South Korea's FSC puts a February 2027 start date on tokenized stocks and bonds

Start with the cap that will shape the products: new fractional offerings carry a subscription limit of 30 million won or 5 percent of the deal's total size, whichever is smaller ^[16], and 30 million divided by 0.05 is 600 million won ^[13], so below that deal size the percentage binds and above it the flat won figure does. Every fractional deal larger than 600 million won is therefore sold in slices no bigger than 30 million, which makes a sponsor's cost of raising a function of how many accounts it can reach rather than how deep any single one of them is.

The venue side runs on the same logic. Retail investors on over-the-counter venues face an annual net-purchase limit of 100 million won per exchange ^[15], so a single platform needs 10,000 retail accounts trading at the cap to clear a trillion won of net retail buying in a year ^[14]. And because the limit is written per exchange rather than per person ^[15], the binding constraint sits on the platform's book, not on the investor's wallet.

Then the cost split, which is where the incumbency question lives. Existing securities firms may handle tokenized securities under their current licences with no new authorisation ^[7], while a non-bank issuer must hold 4 billion won of equity capital and staff dedicated to account management, compliance and IT ^[8]. The brokerage decision is a systems budget; the entrant's decision is 4 billion won and three functions before it issues anything.

On the timetable itself: only the first phase carries a date, February 4, 2027, when the amended securities law takes effect, and it opens with privately placed money-market funds, corporate bonds for institutional investors, unlisted stock issued through a trust structure, and publicly offered fractional-investment securities ^[4]. The second phase widens to all publicly offered securities and the third aims at settlement directly on the blockchain, with stablecoins as the settlement asset ^[5], but the roadmap as published attaches no dates to either ^[6]. Five months separate the September 4, 2026 publication from phase one ^[10]; phases two and three follow it in sequence, with no dates attached to either. The plan came out of the third meeting of a public-private consultative body held at the Korea Securities Depository ^[2], which tells you the incumbent depository is in the room while its eventual replacement is being sketched.

The chip adjacency is worth naming and then setting down. The FSC's stated motive is modernising capital markets and widening the ways companies and investors raise and deploy money ^[9], and it lands while semiconductor exports are up 167.7 percent in May 2026 from a year earlier ^[11] and the Bank of Korea argues the expansion that began in March 2023 has run 40 months, 11 months past the average of the five cycles from 2000 to 2020 ^[12]. The published plan draws none of that chip money. What the windfall buys is political cover for the agenda; the technical work of building it stays where it was.

The plan supports three readings. The first holds that phase one is plumbing: institutional MMFs and corporate bonds get mirrored on-chain, settlement stays where it is, and the retail caps never bind because retail is barely present until phase two. The second holds that the trust-structured unlisted stock channel turns out to be the real product, because that is the instrument with no good existing venue. The third holds that phase three arrives late or arrives bank-only, and the stablecoin leg is a design intention rather than a market. The evidence supports the first reading most cleanly, since the dated phase is the institutional one ^[4]. What would falsify it: a cohort of new non-bank issuers clearing the 4 billion won bar ^[8] before February 2027, or a date attached to phase two.