The Korean ELS Machine: Why Brokerages Sell Structured Products Even After Last Time It Broke

South Korea's KOSPI had just entered a bear market — down nearly 20% from its peak in two weeks, triggering emergency circuit breakers for the first time since the 2008 financial crisis. And what did Korean retail investors do in response? They rushed toward more structured products promising 40% coupon returns.

That is not the part of the story that sounds rational. It is the part that makes the rest of it make sense.

At the center of this machine is NH Investment & Securities, one of Korea's largest brokerages, which has been filing new series of equity-linked securities (ELS) in near-continuous motion. On September 10, it filed for another 10 tranches — the N2 ELS 431st through 440th — worth KRW 57 billion, just one more notch under an existing 10-trillion-won issuance program. Three-year maturity. High-complexity. Principal not guaranteed. The prospectus tells investors to buy only if they can bear a 100% loss. The subscription window opens in a week and runs through September 22.

This is the same firm whose first-half 2026 net profit reached KRW 965.1 billion, a 108% increase year-over-year, fueled by a surge in brokerage commissions and financial product sales. The ELS business is not a side note. It is one of the profit engines.

The headline from a Korean wire service called the demand for the latest N2 issuance "mixed." That word is doing a lot of work. But before asking whether demand is strong or weak, it is worth understanding what investors are actually buying — and what the brokerage is actually earning.

What an ELS is, and who holds the risk

An equity-linked security is not a stock, and it is not a bond, despite the name. It is a structured note — a fixed-income wrapper around a set of option positions. The investor does not own any underlying shares. What the investor holds is a promise from the issuer, plus the payoff profile of what is essentially a short put option on a basket of stocks or an index.

Here is how the payoff typically works. The investor hands over principal. In return, they get an above-market coupon — in Korea, often 8% to 15% annualized on "standard" products, and sometimes much higher on the riskiest tranches — if the underlying assets stay above a predetermined "knock-in" barrier through the life of the product. If one of the underlying names crashes below that barrier, the investor absorbs a loss proportional to the decline. Total loss is possible. The product is cash-settled at maturity, so there is no chance for the stock to recover after the note expires.

The investor is selling a put. The issuer — NH Investment & Securities, in this case — is the counterparty. But NH doesn't typically hold that risk. It hedges the option exposure in the market. The fee income, the underwriting spread, and the structured-product sales commission are what stays on NH's books.

This is basically the same financial machine that blew up in 2024. That year, Korean banks and brokerages had sold over W19.3 trillion of HSCEI-linked ELS. Investors got roughly 5% annually if the index stayed between 50% and 110% of its starting value. When the index crashed — from above 12,000 to around 5,000 — the knock-in barriers were breached and investors lost roughly half their principal, with the loss rate hovering at around 50%. Four major banks alone confirmed investor losses totaling W312.1 billion out of W588.8 billion in products sold. The broader retail loss was in the hundreds of billions more.

The mechanics were identical to what NH is selling today. Different underlyings — now Samsung Electronics, SK hynix, Tesla, Intel, Micron, and the KOSPI 200 — but the same structure: investor sells a put, collector earns fees, and the knock-in barrier is the tripwire that turns a coupon note into a stock-loss product.

Why "mixed demand" is the real headline

The competitor's article used "mixed demand" to describe how different tranches in the latest N2 series performed. The article did not explain what the tranches were, but the pattern in Korean ELS markets is telling. Korean name-linked ELS products gain traction, while products tied to foreign equities — especially volatile ones like Tesla and Intel — or to broad indices tend to draw less enthusiasm. Investors who have been burned once can still see the difference between Samsung and a five-name basket with a knock-in barrier, even if both are technically "high-risk, principal non-guaranteed."

This split in demand is the useful part of the story. It is not that Korean retail investors have somehow amnesia about 2024. It is that they are selectively re-entering the same product line, and the selection itself reveals what they think they are buying.

The low-risk ELS tranches — the ones that offer 99% principal repayment with no knock-in condition and a 6.70% annual pre-tax return — sell out. NH has expanded its retirement pension ELS lineup specifically to capture investors who want ELS-adjacent returns with a loss floor. The highest-yield tranches, the ones offering the most attractive coupons but carrying full principal risk, are where the demand thins.

What this looks like from NH's balance sheet: the fee income is reliable across the whole spectrum. A sold-out conservative tranche generates just as much product-sales commission as a partially subscribed aggressive one. The real question is not whether NH is earning — it clearly is. The question is whether a whole class of Korean retail investors is slowly rebuilding the same risk exposure that nearly broke them last time, just with different underlyings and slightly better product labels.

The regulatory tripwire

Korea's financial regulators are not sitting still. The Financial Supervisory Service, after a reform task force operated from March to June 2026, announced a package of ELS reforms scheduled for September 2026 — the same month NH's latest N2 tranches open for subscription.

The changes are targeted. For high-complexity ELS products, brokerages must now provide advance notice once when the price of an underlying asset approaches the knock-in barrier by 10 percentage points. Self-inspection frequency increases from once a year to once a quarter, and board reporting from once a year to once every half year. Brokerages face new requirements around product design, sales practices, and post-issuance management.

These are guardrails, not a prohibition. The ELS machine is not being dismantled — it is being fitted with warning lights. The economic incentives underneath have not changed. Investors still get coupons for selling downside protection. Brokerages still earn fees for structuring and selling the notes. The knock-in barrier still exists as the line between "nice yield" and "you own a crashed stock economically."

What the rules do change is the cost structure on the brokerage side. More inspections, more disclosures, more documentation. That is a margin pressure that flows through NH's operating results over time, even if the current revenue surge — fee and commission income up 112% year-over-year — makes it invisible right now.

What this means for watching NH Investment & Securities

NH is a well-capitalized firm. Its net capital ratio of 2,952%, with net profit attributable to controlling interests at KRW 965.2 billion and net operating revenue of KRW 2.26 trillion shows the Korean market boom has flowed through to earnings. The stock, though, has fallen from a 52-week high of 42,600 KRW to trade at 26,250 KRW — a reflection of the KOSPI selloff and the foreign money that has been pulling out of Korean equities.

For an outside investor trying to read the business, the ELS machine raises two questions that the quarterly earnings won't answer directly.

First: how much of NH's profit growth is structural — a durable shift in Korean retail investors toward structured products — and how much is cyclical, tied to this specific window where a crashing market is pushing savers toward high coupons instead of out the door entirely? If it is cyclical, the ELS revenue is a boom-bust line item, not a growth story.

Second: what is the hidden liability of all those outstanding ELS products on NH's books? NH hedges its option exposure, but hedging is not costless — and it is most expensive exactly when the underlying assets are crashing, which is when the knock-in barriers get hit and the investor losses materialize. The 2024 China ELS crisis forced banks to absorb hedge losses while their retail clients absorbed principal losses. The next Korean ELS stress event — whether from semiconductors, a geopolitical shock, or another China contagion — will test whether NH's hedging program and capital buffer can handle the tail.

The "mixed demand" headline is actually a decent window into the whole picture. Investors have not forgotten 2024. They just haven't stopped chasing yield either. The ones who buy the conservative tranches are getting a coupon with a floor. The ones who buy the high-risk tranches are getting a coupon and a knock-in barrier, and they know the difference.

NH earns from both groups. The question for anyone watching the stock is whether the ELS business is a steady fee engine or a delayed stress test wearing a fee engine's clothes. The regulatory changes coming this September will add friction to the machine. They won't change what the machine does.