Silicon2 will issue 6,666,666 preferred shares for almost ₩300.0bn. Silicon2 Co., Ltd. (KRX:257720) says the capital will support global infrastructure and market-share expansion. The funding validates ambition, but its redemption rights make the terms as important as the headline amount.
The operating case is strong. First-half revenue rose to ₩749.2bn, and operating income reached ₩147.5bn. The working-capital case is less clean. Inventory climbed to ₩451.0bn, while second-quarter operating cash flow turned sharply negative.
Silicon2 approved a third-party allotment to Starlink Investment L.P. on August 18. A third-party allotment issues new securities to a selected investor instead of offering them pro rata to existing shareholders.
The company will issue 6,666,666 redeemable convertible preferred shares. Total proceeds are ₩299,999,970,000, which the filing rounds naturally to ₩300.0bn. The issue price works out to ₩45,000 per share from those disclosed figures.
These are not ordinary common shares. The preferred stock carries voting rights and can convert into common stock. The initial conversion ratio is one common share for each preferred share.
The preferred holder can also request redemption starting three years after issuance. The redemption amount includes the issue price plus a 2% annual compound return. Redemption remains subject to legally available distributable profit.
That structure creates two possible claims on future value. Conversion can dilute common holders. Redemption can require cash if the holder chooses that route and the legal conditions are met.
The preferred-share count equals roughly one share for every ten common shares outstanding before the deal. That does not mean immediate common dilution because the instrument begins as a separate class. It shows the potential scale.
The shares carry a one-year lockup. This prevents an immediate market sale after issuance. It does not remove later conversion or redemption rights.
The investor also deserves attention. Starlink Investment L.P. was formed in Jersey in May 2026. The filing says it had only ₩141,490 of assets when the transaction was disclosed.
That tiny starting balance sounds alarming without context. The filing also says the partnership will complete capital funding before paying for the shares. Investors still need confirmation that payment occurs as scheduled.
Silicon2 distributes many Korean cosmetics brands rather than depending on one owned label. It operates an e-commerce platform and a wholesale network. The United States is among its largest markets.
This aggregation model has a useful advantage. When one Korean brand goes viral abroad, Silicon2 can capture distribution demand without bearing the full product-development risk. A broad catalog can follow consumer attention.
The financial record shows rapid scale. Revenue increased from ₩64.9bn in 2019 to ₩131.0bn in 2021. It reached ₩342.9bn in 2023 and ₩691.5bn in 2024. Latest annual revenue was ₩1.12tn.
Profit grew with sales. Operating income rose from ₩4.4bn in 2019 to ₩47.8bn in 2023. It reached ₩137.6bn in 2024 and ₩205.4bn in 2025.
The first half of 2026 extended that trend. First-quarter revenue reached ₩346.6bn, up from ₩245.7bn one year earlier. Second-quarter revenue increased to ₩402.6bn from ₩265.3bn.
Operating income was ₩64.5bn in the first quarter and ₩83.0bn in the second. The comparable quarters produced ₩47.7bn and ₩52.2bn.
Net income reached ₩126.7bn for the half. It was ₩74.4bn in the prior-year period. Silicon2 is not sacrificing reported profit to expand.
The placement can accelerate this model. More warehouses and local infrastructure can shorten delivery times and improve brand access. The filing does not identify spending by country, so returns will have to be judged from later reports.
Silicon2 held ₩451.0bn of inventory at the end of June. That was up from ₩300.1bn at the end of 2025. It stood at ₩246.6bn one year earlier.
Inventory expansion is central to the distributor model. The company must stock enough products to serve multiple regions and respond to sudden demand. A broader catalog also requires more working capital.
The risk is obsolescence. Beauty trends can change quickly. Products that once sold through viral channels may require discounting if consumer attention moves.
Receivables also increased. They reached ₩118.7bn in June, compared with ₩59.6bn one year earlier. Stronger wholesale sales may explain part of that rise, but cash collection must follow.
Quarterly cash flow shows the strain. First-quarter operating cash flow was positive ₩105.9bn. It turned negative by ₩92.4bn in the second quarter.
The first-half total remained positive at ₩13.5bn. That was still far below net income of ₩126.7bn. The gap reflects how much cash growth can absorb before customers pay.
Cash stood at ₩82.3bn at the end of June. Current liabilities were ₩221.9bn. The preferred-share proceeds would materially expand liquidity if the transaction closes.
The company also spent ₩31.1bn on property and equipment in the second quarter. First-quarter purchases were only ₩917m. This supports the claim that infrastructure investment is stepping up.
The company's market capitalisation was ₩3.2 trillion at Thursday's close. The shares finished September 3 at ₩48,500. Their annual trading range is ₩30,400 to ₩54,200.
Latest annual net income was ₩168.6bn. The market value is about 19 times that result. First-half 2026 net income of ₩126.7bn suggests the earnings base is still growing.
This historical comparison looks more grounded than the ratios attached to many fast-growing exporters. It does not include the economic claims created by the preferred shares.
The ₩300.0bn proceeds add assets when funded. They also create securities with conversion and redemption rights. Treating the cash as pure common equity would overstate the benefit.
Accounting equity was ₩688.9bn at the end of June. It increased sharply from ₩457.4bn at year-end. The financial context does not explain every part of that change, so I won't attribute it entirely to retained profit.
The valuation assumes Silicon2 can keep gaining from global K-beauty demand. It also assumes warehouses and inventory earn acceptable returns. The new capital makes both outcomes more consequential.
K-beauty demand is not guaranteed to grow in a straight line. Social-media trends can shift. Foreign retailers may change shelf space, and individual brands can lose relevance.
Aggregation reduces dependence on one brand. It does not remove category risk. Silicon2 still needs Korean beauty products to remain attractive across its overseas markets.
Inventory is the clearest measurable risk. It now exceeds half of current assets. A slowdown could force markdowns and weaken gross profit.
Cash conversion is already uneven. The second quarter produced strong earnings and a large operating cash outflow. One quarter can reflect shipment timing, but repeated gaps would signal lower earnings quality.
The Starlink funding has execution risk. The partnership was newly formed and had nominal assets when disclosed. Its planned capital contributions must arrive before Silicon2 receives the money.
The security terms matter after closing. Conversion can increase the common share base. Redemption at a compound return can create a future cash obligation.
The investor's voting rights also mean this is more than passive financing. The supplied filing identifies no prior relationship with the company or controlling shareholder. Future governance disclosures will show how the holder participates.
The next proof should come in two places. First, the company should confirm payment and issuance of the preferred shares. Second, quarterly operating cash flow should catch up with profit as inventory moves to customers.
Silicon2 has earned its growth reputation through revenue and operating income, not slogans. Starlink's ₩300.0bn commitment can fund the next stage. It also raises the cost of poor inventory decisions. The deal works best if faster global expansion produces cash before redemption rights become relevant.
kstock reads DART every morning and writes up what moved — the contract, the buyback, the number that does not add up. The daily post and a Saturday roundup, by email.