Exchange-event research
Upbit caution / warning flag: what happens to the price
Upbit flags a coin with an "investment caution" when it sees a risk — a sharp drop in volume, problems with the team or disclosures, a pending review — and a flag that is not lifted can end in a delisting. Below: how the price moved after every caution notice and every lifted caution we have recorded on Upbit, with Bithumb's equivalent next to it.
Built from 83 measured coin-events since May 2023; updated as new announcements are priced.
Questions and answers
- What happens to the price after Upbit flags a coin with an investment caution?
- Caution / monitoring tag · Upbit: median -1.5% after 1 hour, -3.3% after 24 hours and -8.6% after 7 days; 33% of coins were higher 24 hours later (n = 67).
- What happens to the price when Upbit lifts the caution?
- Caution lifted · Upbit: median +3.7% after 1 hour, +1.8% after 24 hours and -2.5% after 7 days; 50% of coins were higher 24 hours later (n = 10).
Median price move after the announcement
Per exchange: n = coins measured, the median move from the price at the announcement, and the share of coins higher 24 hours later.
| Group | n | Median 1h | Median 24h | Median 7d | Higher after 24h |
|---|---|---|---|---|---|
| Caution / monitoring tag · Upbit | 67 | -1.5% | -3.3% | -8.6% | 33% |
| ↳ from when we caught it | 3 | caught live on time: 3 — too few to summarise yet | |||
| Caution / monitoring tag · Bithumb | 4 | Fewer than 5 measured — too few to summarise. | |||
| ↳ from when we caught it | 2 | caught live on time: 2 — too few to summarise yet | |||
| Caution lifted · Upbitsmall sample | 10 | +3.7% | +1.8% | -2.5% | 50% |
| ↳ from when we caught it | 0 | caught live on time: 0 — too few to summarise yet | |||
| Caution lifted · Bithumb | 2 | Fewer than 5 measured — too few to summarise. | |||
| ↳ from when we caught it | 1 | caught live on time: 1 — too few to summarise yet | |||
Medians of past moves, measured on the same market from the price at the announcement. Past reactions don't predict the next one: listing and delisting moves are fast and often reverse. Not financial advice. The main rows are measured from the price at the official announcement time, which is usually not tradeable on a fast listing. The “from when we caught it” rows use only announcements our feed stored within 15 minutes, measured from the first price after that.
Latest examples
The newest events of this type and the move 24 hours later. Open one for the full record.
| Published | Announcement | Coin | +24 hours |
|---|---|---|---|
| Oct 16, 2026 | Upbit · Caution / monitoring tag만트라(MANTRA) 거래 유의 종목 지정 기간 연장 안내 | MANTRA | -4.1% from when we caught it |
| Oct 13, 2026 | Upbit · Caution / monitoring tag블러(BLUR) 거래 유의 종목 지정 안내 | BLUR | -11.4% from when we caught it |
| Oct 2, 2026 | Upbit · Caution / monitoring tag블라스트(BLAST) 거래 유의 종목 지정 안내 | BLAST | -35.5% from when we caught it |
| Oct 2, 2026 | Upbit · Caution lifted샌드박스(SAND) 거래 유의 종목 지정 해제 안내 | SAND | +62.3% caught late (2 hr) |
| Sep 22, 2026 | Upbit · Caution lifted인젝티브(INJ) 거래 유의 종목 지정 해제 안내 | INJ | +5.2% caught late (242 hr) |
| Sep 22, 2026 | Upbit · Caution / monitoring tag소폰(SOPH) 거래 유의 종목 지정 안내 | SOPH | -14.1% caught late (239.4 hr) |
| Sep 21, 2026 | Upbit · Caution / monitoring tag멀티버스엑스(EGLD) 거래 유의 종목 지정 안내 | EGLD | +4.7% caught late (262.4 hr) |
| Sep 18, 2026 | Upbit · Caution / monitoring tag만트라(MANTRA) 거래 유의 종목 지정 기간 연장 안내 | MANTRA | +3.9% caught late (334.4 hr) |
| Sep 18, 2026 | Upbit · Caution / monitoring tag질리카(ZIL) 거래 유의 종목 지정 기간 연장 안내 | ZIL | +5.1% caught late (334.4 hr) |
| Sep 1, 2026 | Upbit · Caution / monitoring tag인젝티브(INJ) 거래 유의 종목 지정 안내 | INJ | -0.7% archive |
How this is measured
Every announcement from the exchange's public feed is stored with the coin's price on the same market (Binance or Bybit, spot first) at the moment it was published, then 1 hour, 24 hours and 7 days later. Brand-new tokens whose trading started after the announcement are left out of the medians: their first price is an opening auction, not a reaction to the news. One coin-event is one ticker in one announcement; moves beyond ±1000% are treated as bad data and left out, and groups of fewer than 5 coins show no median. Older announcements come from the exchanges' archives and are priced from historical candles.
More exchange-event research
How the price moved after past announcements doesn't predict the next one. Listing and delisting moves are fast and often reverse. Not financial advice.