Invest1 distinct publisher3 min readUpdated
The bank says it reconciled repayment records against deposits every day. Both sides of that check came from the counterparty now accused of diverting 83.376 billion won.
The Investor · Invest desk

Compiled by The InvestorSomething wrong?How this is made
Put the bank's own figure against the regulator's own dates and you get a daily rate. The Financial Supervisory Service times the fraud from December 1 to June 29 [10], a span of 211 days [18]. IBK's disclosure last month put the total at 83.376 billion won [12]. That is roughly 395 million won a day of borrower principal and interest going somewhere other than the bank [19], for most of two fiscal halves.
The reason a daily check could not see it is structural. IBK disbursed the loans itself, but repayments travelled through an account designated by Company B, the online lending platform, and were settled with the bank later [4]. Company B changed that account and kept the money [5], then processed false computer records showing repayments that had never been transferred [6]. IBK says it compared repayment records against actual deposits every day [8]. Both sides of that comparison originated with the same counterparty, so the check could clear every morning and prove nothing. What actually broke the case was settlement money failing to arrive on June 24, plus rising borrower complaints [9], about 206 days after the diversion began [20]. The FSS window runs five days past the date IBK says it noticed [22].
A cheaper signal was on offer earlier. Five Chinese financial institutions removed Company B from their partner lists in March and April, and IBK did not pick it up [11]. Taking the later of those two months, that leaves at least eight weeks in which the platform collecting IBK's loan repayments was being dropped by local peers while IBK kept lending through it [21].
The remedy is the most telling detail. IBK has now built a separate system letting borrowers check repayment amounts or repay early without going through Company B [14]. That is not sophisticated engineering. It is the arrangement the product needed on day one, and building it now concedes the design was the exposure. In the meantime, borrowers who had cleared their loans were recorded as in arrears or default, pursued by collections and marked down on their credit scores, while IBK failed to recover the principal and interest [7].
Two months after becoming aware, the bank still cannot state the actual amount involved, the sum recovered or the expected loss, and points to a local investigative agency [13]. The supervisor's posture matches. After the case was reported, the FSS filed internal reports, circulated it and checked whether the amount had changed, but did not begin face-to-face verification with IBK officials, request further materials, or open a written probe or on-site inspection [15]. It says it will review the appropriateness of the handling once IBK submits a closing report after cleanup, including recovery of the losses [16]. According to Rep. Shin Dong-wook of the People Power Party, who obtained the materials from the FSS and IBK [2], IBK "handed out the loans directly but left the collection of principal and interest to an outside platform," and the FSS's "belated oversight, simply waiting for a closing report," is a serious problem in its own right [17].
For any lender running outsourced origination or collection rails, the transferable lesson is narrow and unglamorous: the party that holds the account of record cannot also be the party that supplies the record.
Follow any of these and your For You feed starts watching them — no settings page required.
Ranked by verification strength, evidence, and original report placement.
Weak internal controls worsened a financial fraud worth more than 80 billion won ($60 million) at the Chinese unit of Industrial Bank of Korea, a lawmaker said.
Materials on the case were submitted by the Financial Supervisory Service and IBK to Rep. Shin Dong-wook of the People Power Party, a member of the National Assembly's National Policy Committee.
IBK's Chinese unit signed an agreement with a local non-bank financial firm, referred to as Company A, to make contactless loans to borrowers in China; Company A handled borrower recruitment and principal-and-interest repayments through an online lending platform, Company B.
IBK disbursed the loans directly, but principal and interest repaid by borrowers passed through an account designated by Company B and were later settled with IBK.
Company B arbitrarily changed the repayment account and diverted the principal and interest paid by borrowers instead of sending it to IBK.
An investigation found that Company B processed false computer records making it appear that repayments had been made when no money was actually transferred.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Documented by regulator and bank submissions, but single-outlet and incomplete
The factual spine comes from materials the FSS and IBK themselves submitted to a National Assembly committee member, which is stronger than anonymous sourcing, and it includes specific dates, a disclosed amount and named supervisory steps not taken. It is nonetheless one publisher's account, the investigation is still with a local investigative agency, and the bank's own loss numbers are explicitly unsettled, so the record supports the structure of the failure more firmly than its final magnitude.
Concrete but narrow: one disclosure, one remediation deployment, no supervisory action
Observable real-world action is limited to a single institution: IBK's mid-July disclosure of 83.376 billion won and its post-incident direct repayment channel that bypasses Company B. The counterparty-risk signal was acted on by five Chinese institutions delisting Company B, but IBK did not follow, and the supervisor has taken no verification or inspection step. There is no evidence of sector-wide policy change, broader remediation, or usage metrics for the new repayment system.
Slightly understated: the headline loss is a floor, not a settled figure
The story's quantitative claims are anchored to the bank's own disclosure and the FSS window rather than inflated: the article itself notes that the actual amount, recoveries and expected losses remain undetermined, and the FSS dates the fraud five days past the bank's detection date. Framing language such as 'weak internal controls' is attributed to a lawmaker, and the concrete mechanism (one platform holding both the cash and the ledger) is documented. The direction of residual error therefore leans toward understatement of exposure, though only modestly, because borrower-count and financial-impact detail is absent.
Oversight-driven sourcing with self-interested framing on all named sides
Every substantive input carries a stake. The material reached the public through a lawmaker who is publicly criticising both the bank and the regulator, giving an accountability incentive to emphasise control failure. IBK has a clear interest in framing the episode as external fraud and in asserting a daily reconciliation control that its own timeline undercuts. The FSS has an interest in deferring judgement until a closing report arrives. No neutral investigative finding is yet available to arbitrate.
Mechanism well evidenced; magnitude and consequences still open
Confidence is moderate. The structural facts (who disbursed, who collected, who controlled the records, when the window ran, what the supervisor did not do) are supported by documents from both the bank and its regulator and are internally consistent. Confidence is held back by single-publisher coverage, an open criminal or administrative investigation, unresolved loss magnitude, and the absence of any statement from the accused platform or from other institutions with similar arrangements.
invest
Korea's bankruptcy count hits 1,500 in seven months, and a 27% slowdown still sets a record1 distinct publisher
invest
The FSS relocation plan's real bill: 85.6% of staff would consider quitting1 distinct publisher
invest
Regulators head for Sejong, and the access bill lands on the firms they supervise1 distinct publisher
invest
Korea's 3-million-won emergency loans jump 39.2% while banks squeeze everything else1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
en.sedaily.com
1 article · August 23, 2026