Live · updated 2 minutes ago · 119 stories in 24h

← Stream
Stocks & Indices Unverified First seen

Elice IPO plunges on debut, deepening AI valuation concerns

Hedgeweek reports that South Korean AI infrastructure provider Elice plunged 32% on its trading debut, closing at 61,500 won after falling as much as 38%, the worst opening-day performance for a major South Korean IPO on record. CoinDesk reports that South Korea's KOSPI fell almost 4% alongside the weak debut, with Elice tumbling around 30%. CoinDesk also links the debut slump to concerns over AI infrastructure valuations and notes rising oil prices compounded market uncertainty.

Why “Unverified”

  • 2 independent sources (Hedgeweek, CoinDesk), but no primary confirmation

2 independent source groups across 3 posts · evidence score 79/100. How labels work

What the sources claim

  1. Elice shares fell 32% on their trading debut, closing at 61,500 won after dropping as much as 38%.

    “Shares in South Korean AI infrastructure provider Elice plunged 32% on their trading debut on Tuesday”
    Hedgeweek
  2. The Elice debut was the worst opening-day performance for a major South Korean IPO on record.

    “marking the worst opening-day performance for a major South Korean IPO on record”
    Hedgeweek
  3. South Korea's KOSPI fell almost 4% alongside Elice's weak debut.

    “South Korea’s KOSPI fell almost 4%, accompanied by one of the weakest IPO debuts in the country’s history”
    CoinDesk
  4. Elice tumbled around 30% in its debut, described as one of the weakest in South Korea's history.

    “with data centre company Elice tumbling around 30%”
    CoinDesk
  5. Elice's debut slump added to concerns over AI infrastructure valuations.

    “Elice’s debut slump added to concerns over AI infrastructure valuations”
    CoinDesk
  6. Rising oil prices compounded market uncertainty.

    “while rising oil prices compounded market uncertainty”
    CoinDesk

Sources

More in Stocks & Indices

All Stocks & Indices stories →

Summary and claims extracted by AI with prompt analyze-story@5 on 2026-10-20 10:20 UTC. Labels are computed from the sources, not by the model.