As of October 29, 2026, Chow Tai Fook reported retail sales down 0.5 per cent for the three months to September 30, a quarter in which mainland China fell 5.3 per cent while Hong Kong, Macau and other markets rose 33.2 per cent, and in which the group kept closing stores. The jeweller also said net profit for the six months to September is expected to rise by about 90 to 100 per cent. Flat revenue, fewer stores and a much higher profit line is the shape of a network being rebuilt around ticket size rather than coverage.
What the quarter actually shows
The mix shift is the story inside the flat number. Mainland China accounted for about 87.5 per cent of group retail sales value a year earlier and 83.3 per cent now, a move of roughly four percentage points in twelve months. That is not a rounding error; it reflects a market where demand is soft and a second market where tourism and gold buying are strong, and the two now pull the average in opposite directions.
The profit guidance needs reading with the same care. A 90 to 100 per cent rise in net profit is a large number, but the company attributes part of it to the gold price. Store economics and metal prices move on different clocks, and a result flattered by the gold price does not repeat once the price flattens, which is why the operating detail matters more than the profit headline.
Fewer stores, bigger tickets
The Dawn line sold more than HK$870 million in its first three months, and the company has been pushing design-led products while closing stores. Those two decisions support each other: a signature collection needs display space and trained staff, and a store that cannot present it properly is better closed than kept open as a distribution point for generic stock.
Why a signature line changes site selection
When a product needs explaining, the store becomes a showroom rather than a warehouse. That pushes site selection towards fewer locations with higher footfall and better adjacencies, and away from coverage for its own sake. A high-potential store framework captures the same logic: the question is not how many people pass the door but how many of them are there to buy jewellery.
The store network is being reshaped, not shrunk
Closing stores is the visible half of the change. The other half is where the remaining stores sit and what they carry. A network built for coverage puts a small counter in every mall; a network built for ticket size puts a larger format in the malls where the customer already shops for other categories, and accepts that some towns will be served online instead.
| Network decision | What it changes | Metric to watch |
|---|---|---|
| Close underperforming stores | Lowers coverage, lifts sales per store | Sales per square foot on retained sites |
| Enlarge flagship formats | Adds display space for design lines | Average ticket and conversion |
| Shift mix by market | Weights the estate towards stronger markets | Revenue share by region |
| Push signature collections | Raises the reason to visit | Sell-through of new lines |
The table also shows why store count is a poor headline. Two retailers with the same number of doors can have very different economics, depending on whether those doors are in trade areas that support the range. For a jeweller moving towards design-led products, a smaller estate in better locations is a growth strategy, not a retreat, and the numbers in this quarter are consistent with that reading.
What a trade area test would show
A trade area test would ask how far customers travel for a purchase of this size, and what else sits within that radius. Jewellery buying is infrequent and considered, so the relevant catchment is wider than for groceries but narrower than for furniture. Stores outside that catchment rarely recover the cost of holding stock, however good the local footfall looks.
Auditing a store network during rationalisation
During a store rationalisation, the audit has to answer two questions at once: which sites earn their rent, and which products justify the space they occupy. The sequence below starts with the site, moves to the range, and ends with the market mix, so that closures and investments can be decided in the same review rather than in separate budget cycles.
The output of this audit is a portfolio decision rather than a list of weak stores. Some sites are worth enlarging because the trade area supports a bigger format; others should close because no format will earn its rent there. Treating every underperforming store the same way is how retailers end up closing the good ones and keeping the ones with the longest leases.
FAQ
Why close stores when sales are only slightly down?
Because a flat sales line hides a change in where the money comes from. If a smaller estate produces the same revenue, the closed stores were not paying their way. Closing them lifts sales per store and releases rent, which matters more to profit than the headline revenue figure suggests.
What does a design-led line change for a store network?
It changes the role of the store. A signature collection needs display space, trained staff and a reason for customers to visit rather than buy online. That pushes the estate towards fewer, larger sites in strong trade areas, which is the opposite of a network built for coverage.
How much of the quarter came from gold prices?
The company said profit for the six months to September is expected to rise by about 90 to 100 per cent, and part of that reflects the gold price rather than store performance. Separating the two matters, because a price-driven gain does not repeat once the metal price flattens.
Which store metrics should a jeweller track?
Sales per square foot, average ticket and the share of sales from design-led lines. Coverage metrics such as store count matter less once a network is being rationalised, because adding stores in weak trade areas dilutes the average rather than growing the business.
