Editor’s note: This is general educational information about how IPO price discovery is regulated in Korea. It is not investment advice, and every rule and figure below comes from the official sources listed at the end.
Book-building is often described as the market deciding what a company is worth. In Korea it is closer to a regulated sequence with defined participants, defined disclosure moments and, from late 2026, a defined pre-marketing stage that used to be illegal. The Financial Services Commission has rebuilt most of that sequence since December 2022, and the changes are specific enough that an investor reading an IPO can tell which version of the process a deal ran through.
The Problem the Rules Were Written For
When the FSC set out measures to improve the soundness of the IPO market on December 19, 2022, it named three faults. Setting a price band was difficult because there was no reliable way to see how demand shifted as the price moved. At the subscription stage, institutions bid for volumes far above their real demand in order to secure a larger allocation, a practice the FSC called fictitious oversubscription, which fed excessive competition and more oversubscription. And after listing, quotation prices repeatedly hit the daily upper limit on the first day, or the first and second days, producing what the FSC described as a de facto transaction halt followed by a collapse.
The remedies attacked each fault at its own point in the chain. Bookrunners were to be allowed to test demand before filing a securities registration statement. The book building period, then completed in two days, was to be extended, for example to seven days. And the quotation price range on the day of listing was raised to 60 to 400 percent of the public offering price, chosen after a review of practice in other countries, so that an equilibrium price could form on the first day instead of being suppressed by a limit.
Making the Bid Mean Something
The anti-fictitious-bidding rule arrived on April 26, 2023, when the FSC approved a revision to the regulation on financial investment businesses at its eighth regular meeting. The bookrunner became responsible for verifying an institutional investor’s ability to pay before shares are allotted, with an administrative fine for unfair transaction activity where it fails to do so, and the requirement applied to securities registration reports filed after July that year. Bookrunners also gained the power to penalise institutions caught bidding beyond their capacity, through reduced allocations or exclusion from book building, and allocation was restricted for institutions that failed to state a bidding price at the book building stage.
That last provision is the heart of the reform. An order book in which some participants name a quantity but no price is not a demand curve, and the underwriter reading it is guessing. Requiring a price on every bid, and requiring the bookrunner to check that the bid could be paid for, is what turns the exercise into price discovery rather than a queue.
Similar plumbing exists in other markets. FINRA Rule 5131, which governs new issue allocations in the United States, requires that a lock-up agreement covering officers and directors also apply to their issuer-directed shares, and that at least two business days before any release or waiver of such a lock-up the book-running lead manager notify the issuer and announce the release through a major news service. The concern is the same one the FSC identified: an allocation process is only credible if the commitments attached to it are visible and enforced.
Preliminary Book Building and the Cornerstone Investor
The two measures the FSC could not implement by regulation alone required primary legislation, and the National Assembly passed the revision bill to the Financial Investment Services and Capital Markets Act on April 23, 2026. Under the existing law, soliciting subscriptions before a securities registration statement is filed is prohibited, so a bookrunner who surveyed institutional demand while setting an initial price band was arguably breaking the law. The revised FSCMA grants an exemption for preliminary demand forecasting, and a second exemption for the cornerstone investor system.
On July 30 the FSC published the implementing detail in a revision proposal for the Enforcement Decree and the regulation on the issuance of securities and disclosure, with the revised FSCMA expected to take effect from November 13, 2026. Institutions qualified to take part in preliminary book building must hold at least KRW30 billion in entrusted assets, in the case of private equity funds and discretionary investment business entities, and must have a minimum level of capacity for corporate valuation and for managing undisclosed information internally. The lead manager may share information that will later appear in the securities registration statement only after preparing an IPO due diligence report, must sign a confidentiality agreement with each counterpart, and must keep a record of the time, the counterpart and the content of what was provided. The FSC warns that helping a third party use that information in breach of the agreement may amount to use of material nonpublic information.
Cornerstone investors face a further test. They must hold equity capital or entrusted assets at least twenty times the value of the IPO shares they intend to subscribe for, a relative rather than absolute threshold chosen so that smaller institutions are not excluded from deals of every size. Their lockup is staggered: six months for 50 percent of allocated shares, eight months for 30 percent, and ten months for 20 percent, which the FSC says is designed to stop sales concentrating on a single day.
The allocation caps differ by board. At present 50 percent of KOSPI IPO shares and 15 to 35 percent of KOSDAQ IPO shares go to institutional investors, excluding high-yield funds and KOSDAQ venture funds. Within that institutional tranche, cornerstone investors may take up to 20 percent in total and up to 10 percent each in a KOSPI IPO, and up to 30 percent in total and up to 20 percent each in a KOSDAQ IPO. Measured against the entire offering, that works out to up to 10 percent in total and up to 5 percent for each cornerstone investor on KOSPI, and up to 4.5 to 10.5 percent in total and up to 3 to 7 percent each on KOSDAQ.
Analysis: What the New Sequence Changes and What It Leaves Alone
The reforms move information earlier and lock shares up longer, and those are two different bets. Preliminary book building assumes the price band was the weak link, because it was set before anyone had legally tested demand. Cornerstone allocation assumes the weak link was the first weeks of trading, when unlocked institutional shares meet a limit-constrained order book. Both can be true, and each is now addressed by a different rule.
The design choices show what each rule acts on. Staggering the cornerstone lockup across six, eight and ten months rather than fixing one date is an explicit statement that a single cliff creates its own selling event. Setting the cornerstone size test at twenty times intended subscription rather than at a flat asset figure keeps small institutions eligible in small deals, which matters more on KOSDAQ, where the higher caps of 30 percent in total and 20 percent per investor are deliberately more generous than the KOSPI equivalents because the deals are smaller.
What the package does not do is remove the underwriter’s discretion over allocation. The order book is still read rather than executed. The rules police who may bid, whether the bid names a price, whether the bidder can pay, and how long the resulting shares must be held, but the final offer price and the distribution of shares remain a judgment made between issuer and lead manager.
For anyone reading a Korean IPO from November 13, 2026 onward, three documents carry the new information. The securities registration statement will follow a demand survey that legally preceded it. The disclosure of cornerstone participation will show how much of the institutional tranche was committed in advance and under which staggered lockup. And the first day of trading will run inside the 60 to 400 percent quotation range rather than against a tight limit, which means the opening print is a market price rather than an artefact of the price limit.
