The average necklace price fell 30.7% and nobody cut a price

Korea's fashion jewelry market grew 22.1% in 2025 while the average purchase price fell 30.7%. But department-store high jewelry rose 60% the same year. What fell was not the price — it was the average.

🇰🇷 Korea·August 16, 2026·9 min read

A market where the average price drops by more than 30% is supposed to shrink. Korea's fashion jewelry market did the opposite in 2025 — it grew 22.1%.

The average purchase price was 43,955 won, down 30.7% from the year before. But in that same year department stores sold 60% more high jewelry, and not one shop announced a price cut.

So the first thing to settle is what actually went down. It was not the price. It was the average.

At a glance

Item2025Change
Domestic fashion jewelry market (imports excluded)702.6 billion won+22.1%
Share who bought in the past year21.1%+7.8pp · all-time high
Buyersabout 9.21 millionabout +3.36 million
Average pieces bought1.84up (women 1.90 · men 1.54)
Average price paid43,955 won−30.7%
Bought at non-brand stores82.4%online 21.2%, a record

The figures come from the Wolgok Jewelry Industry Institute, which commissioned Gallup Korea to survey 1,500 adults nationwide.

One column is missing from that table: how much one person spent. Divide the market by the number of buyers and 2025 works out to roughly 76,000 won per person. Run the same division on the previous year, reconstructed from the published growth rates, and you get about 98,000 won. The number of buyers rose 57% while spending per buyer went down. All six rows point at that one line, and the survey never writes it out.

The market grew because more people bought, not because things got cheaper

Three numbers moved at once, so they can be pulled apart.

What2024 (reconstructed)2025Direction
Buyersabout 5.85 million9.21 million+57%
Pieces per buyerabout 1.551.84+19%
Price per pieceabout 63,400 won43,955 won−31%
Spend per buyerabout 98,000 wonabout 76,000 won−22%

The 2024 column is worked backwards from the published percentage changes, not a figure the institute released. The direction holds anyway: of the four rows, only the first one grew the market. Price per piece fell, spending per buyer fell, and the total still rose 22% because the crowd got 57% bigger.

That also answers why the average dropped. It is not that regular buyers closed their wallets. It is that the 3.36 million newcomers all lined up at the bottom shelf.

Put numbers on it. Suppose the original 5.85 million buyers spent exactly what they spent last year, around 63,000 won a piece, and the 3.36 million newcomers each bought something for 10,000 won. The blended average lands near 44,000 won — almost exactly the published figure. Nobody discounted anything, and the average still fell 30%.

An average measures the mix of people, not the size of a price.

Korean hair salons show the same illusion. The average haircut costs 19,558 won while the "director's cut" costs 60,000. Imagine a counter from an average alone and you will picture a counter that does not exist.

The same year, department-store high jewelry grew 60%

But the bottom shelf was not the only one growing. Shinsegae's high jewelry sales jumped 60% year on year in 2025, and Lotte and Hyundai both rose more than 50%. "People traded down because prices scared them" does not account for that line.

What collapsed was the middle. E-World, a mid-tier Korean jeweler, saw revenue fall 18.2% while operating losses grew 640%. Golden Dew is estimated to have lost about 15% of its revenue. Both sat in the neither-cheap-nor-expensive slot.

So what happened in 2025 was not that jewelry got cheaper. Both ends grew and the middle dropped out.

An average folds both ends into one number, which means it can never show you that shape. The same trap shows up across Korean industry metrics — why Korea counts box office in admissions rather than money is ultimately a choice about what gets folded together.

When gold gets expensive, demand changes material instead of disappearing

So where did the 3.36 million newcomers come from? The age breakdown gives half the answer.

Age groupBought in the past year
20s42.5%
30s34.8%
40s22.3%
50 and over10.9%
All women34.3%

Every band is at or near a record, and two in five Koreans in their twenties bought fashion jewelry in the past year.

Line that up with the 43,955-won average and the counter comes into focus. This is not the world of wedding gifts or anniversary presents. It is the world of something chosen to match what you are wearing today.

The institute's own explanation runs the same way: with high interest rates and high prices, demand for expensive gold-centered jewelry slowed while affordable fashion pieces filled the gap. The important part is that the demand does not vanish. When gold goes up, the person who wanted a gold necklace does not stop shopping — they change material. Silver 925, plating, stainless steel and cubic zirconia take the slot. The impression in a photo is similar; the price loses a digit.

Where that counter sits is also in the data. 82.4% of purchases happened at non-brand stores, and online hit a record 21.2%. You open an app to buy clothes and add a necklace on the way out — and the transaction map of Korea's fashion apps shows those apps expanding in exactly that direction.

What you check at a 44,000-won counter is not a KC mark

But there is one more thing to check here, and it is not the material name. This is the most useful paragraph in this article.

In Korea, metal jewelry that touches skin falls under the Electrical Appliances and Consumer Products Safety Control Act as a "safety-standard-compliance" consumer product. The name is long; the meaning is simple. The government sets limits, but the category is not one that goes through certification and carries a KC mark. So a 44,000-won necklace with no KC mark is normal. What should be there instead is the labeling: material, age of use, the name of the maker or importer, and handling warnings.

The limits themselves come from Korean Agency for Technology and Standards Notice 2018-195.

SubstanceLimit
Nickel release0.5 µg/cm²/week or less
Lead contentunder 0.06%
Cadmium contentunder 0.1%

In March 2022 the Korea Consumer Agency bought 15 pairs of earrings and 15 necklaces from open marketplaces and tested them. Eleven of the thirty exceeded a limit — nickel by 2 to 37 times, lead by 17 to 58 times, cadmium by up to 970 times.

The line after that is the one worth remembering. All thirty failed the labeling requirement. No age of use, no maker's name, no warnings. Some of the products that failed the chemical tests had been advertised as hypoallergenic.

Which means "check the material label" is advice that only works at counters where labels exist. Korean cosmetics have this sorted out far better, and the habit transfers directly — five lines on a label that answer for the seller. A product with no label at all is not cheap. It is missing information.

What this means where you are

Where you areHow you usually reach this counterOne more thing to check
Vietnam · Philippines · IndonesiaMyeongdong and Hongdae street shops on a Korea trip, or Korean fashion apps shipping outIn humid climates plating wears off faster than it does in Korea. For something worn daily, choose silver 925 or stainless steel
ThailandThe same, plus Korean-brand select shops in Bangkok"Hypoallergenic" is ad copy, not a test result. Look for the maker's or importer's name instead
JapanKorean brand pop-ups and cross-border onlineJapan's own nickel rules for jewelry are not the same as Korea's, and this article did not verify them
Anywhere, buying resaleTikTok Shop, Shopee, secondhand appsResale strips the original packaging, and the labeling goes with it

One piece of paperwork applies to all of these. Buy in Korea and you can get the 10% VAT back, but a single receipt has to total 30,000 won or more. One 43,955-won necklace clears that line on its own; five 5,000-won items rung up separately clear nothing. You need your passport, and the shop has to display a TAX FREE sign.

There is one counter where that threshold genuinely bites — in a store where every garment costs 5,000 won, how many times you split the checkout decides whether you get a refund at all.

One thing this article cannot answer: whether the 3.36 million newcomers will still be here next year. People who enter at the bottom sometimes move up and sometimes just stop. What is certain is that the year they leave, the average will swing hard again — and the headline that year will read "jewelry got expensive."

Read next

Sources

Read next