KEPCO Seeks 25 Trillion Won in Prepaid Electricity Fees from Samsung and SK Hynix
nashnova research
Korea Electric Power Corp. is asking Samsung Electronics and SK Hynix to prepay a combined ₩25 trillion in electricity bills to fund grid expansion for semiconductor clusters — a sum equal to roughly a quarter of KEPCO's annual revenue, effectively converting future power bills into upfront infrastructure capital.
Who pays what, and why now?
Samsung is being asked for ₩20 trillion, SK Hynix for ₩5 trillion, according to the Chosun Ilbo; Bloomberg confirmed talks are under way.
The trigger is massive new power demand from chip clusters: the Yongin cluster's early completion plus the new Honam cluster add an estimated 20.6 GW, with nearby AI data centers needing another 7.9 GW.
This means → KEPCO had already earmarked ₩72.8 trillion for grid build-out through 2038, but the scope keeps growing and the money is falling short.
How tight is KEPCO's own balance sheet?
First-half consolidated operating profit was ₩4.91 trillion, but debt had climbed to ₩210.7 trillion by end-June; interest alone cost ₩2.1 trillion in H1.
The harder deadline: a temporary increase in KEPCO's bond-issuance ceiling expires at end-2027. From 2028 the cap reverts to the standard formula.
In plain terms = KEPCO can borrow more right now, but that window closes in roughly two and a half years. If the grid isn't built and the extra borrowing capacity is gone, the projects stall.
How would the prepayment work?
Under the reported structure, the two companies would pay in monthly installments of ₩2 trillion combined over about one year. KEPCO would then deduct electricity charges from the balance each month and pay interest every six months.
The rate under review sits above the two-year Korean government bond yield and at or slightly below the two-year KEPCO bond yield.
This means → KEPCO's funding cost would be lower than issuing its own bonds, while the chipmakers earn a return above sovereign debt and reduce the risk of grid delays at their new fabs — both sides have a reason to agree.
Would this immediately boost KEPCO's profits?
No. Prepayments can only be recognized as revenue when power is actually delivered, so KEPCO's top and bottom lines would not rise right away.
The real effect is pulling forward future electricity receipts to fund grid construction now, shrinking the volume of bonds KEPCO needs to issue.
In plain terms = KEPCO is not earning extra money — it is spending tomorrow's money today, and the payoff is less debt and less interest.
How big is ₩25 trillion in historical context?
Prepaid electricity is not new in Korea, but last year's total was only about ₩390 billion, mostly from government agencies.
The proposed ₩25 trillion is roughly 64 times last year's prepayment volume.
This reflects KEPCO's attempt to transform a small payment-convenience tool into a full-scale grid-investment financing mechanism for major power users — whether it materializes depends on the final rate, term, and both sides' willingness to commit.
市场有风险，内容仅供研究参考，不构成投资建议。