Interest on a Korean margin loan (sinyong georae yungja, a broker-funded stock purchase loan) is not charged by the day you actually borrow. Hold the loan for 45 days and many brokers apply the rate for the entire "31-60 day" tier retroactively across all 45 days. That method is called sogeup-beop, or retroactive application.
On a 30 million won loan held 45 days, the retroactive method produces 318,082 won of interest. The tiered method — checha-beop, which slices the period at each tier boundary and adds the pieces up — produces 292,356 won. Same principal, same days, 25,726 won apart. Which method applies is fixed when the account is opened, and the disclosure usually mentions it in a single line.
What shakes an account before interest does is the maintenance margin ratio. The standard disclosure published by the Korea Financial Investment Association illustrates the product with 4 million won of the investor's own cash plus 6 million won borrowed, buying 10 million won of stock. That account touches the 140% maintenance line after a 16% price drop — not a halving, not a 30% crash.
Why 45 identical days cost 25,726 won apart
Broker rate cards are built as steps. As of 31 August 2026, Mirae Asset Securities quotes general margin lending between 5.90% and 9.50% per year, structured as a 2.89% base rate plus a 3.01-6.61% spread. Kyobo Securities steps differently: 4.4% up to 7 days, 6.5% for 8-15 days, 7.5% for 16-30 days, 8.5% for 31-60 days, 9.2% for 61-90 days and 9.4% beyond 91 days.
The question is how you walk those steps. The standard disclosure lists three conventions. The tiered method cuts the holding period at each boundary and applies that tier's rate to each slice. The retroactive method takes the single rate of the tier the final holding period falls into and applies it to the whole period. The flat method uses one rate regardless of duration. The names sound like a technicality, but the steps span 4% to 9%, so the outcomes do not.
Taking the Mirae Asset tiers (5.90% to 7 days, 7.80% to 15, 8.20% to 30, 8.60% to 60) and running 30 million won for 45 days gives this.
| Holding tier | Days | Annual rate (%) | Tiered method (KRW) | Retroactive method (KRW) |
|---|---|---|---|---|
| Days 1-7 | 7 | 5.90 | 33,945 | 49,479 |
| Days 8-15 | 8 | 7.80 | 51,288 | 56,548 |
| Days 16-30 | 15 | 8.20 | 101,096 | 106,027 |
| Days 31-45 | 15 | 8.60 | 106,027 | 106,027 |
| Total | 45 | — | 292,356 | 318,082 |
The retroactive column applies 8.60% — the rate of the "up to 60 days" tier that day 45 falls into — to all 45 days, so daily interest is a flat 7,068 won and even the first week is priced at 8.60%. The 25,726 won gap is 8.8% more than the tiered result, and on 30 million won of principal it works out to 0.70 percentage points of extra annual cost.
The real trap sits at the tier boundary. Repay on day 60 at 8.60% and interest is 424,110 won. Slip one day to 61 and the 9.20% rate for the 61-90 day tier applies retroactively to all 61 days, producing 461,260 won. One extra day costs 37,150 won — roughly 4.9 times the 7,562 won a single day should cost. For an account that rolls the loan over as a habit, how many of these steps were crossed determines the real cost of funding.
What price drop reaches the 140% line
The margin ratio is account value divided by the loan balance. If you call the loan's share of the purchase amount the leverage ratio, the price drop that reaches 140% collapses into one expression: 1 − (leverage ratio × 1.4). What matters is not how much cash you put in, but what fraction of the position is borrowed.
Holding own capital fixed at 10 million won and varying only the leverage ratio:
| Leverage ratio (%) | Position size (10k KRW) | Loan (10k KRW) | Starting margin ratio (%) | Drop to 140% (%) | Drop to 130% (%) |
|---|---|---|---|---|---|
| 40 | 1,667 | 667 | 250.0 | -44.0 | -48.0 |
| 50 | 2,000 | 1,000 | 200.0 | -30.0 | -35.0 |
| 60 | 2,500 | 1,500 | 166.7 | -16.0 | -22.0 |
| 70 | 3,333 | 2,333 | 142.9 | -2.0 | -9.0 |
A 40% and a 70% leverage ratio produce completely different accounts out of the same stock and the same decline. The first needs a 44% drop to reach the maintenance line; the second needs 2%. Its starting margin ratio is 142.9%, meaning it opens 2.9 percentage points above the threshold. The standard disclosure's example — 4 million won of cash plus 6 million borrowed — is a 60% leverage ratio, which is where the 16% figure comes from.
What shakes an account is not the size of the drop but the share that is borrowed. After the same 20% decline, a 40% leverage ratio still shows 200%; a 70% ratio has fallen to 114.3%, well past the forced-liquidation zone.
Forced sales go out as limit-down orders
A margin shortfall does not trigger a sale the same day. Kyobo Securities asks for a top-up by the next business day when the shortfall occurs with the ratio between 130% and 140%, and liquidates on the second business day if nothing arrives. Below 130% that window shrinks to one business day. The KOFIA standard disclosure similarly states that failure to post additional collateral by the business day following the demand results in discretionary disposal the day after. The gap is a day or two, but the point is that breaking 130% halves the time available.
The disposal price matters more. The order can be placed at the daily limit-down price relative to the previous close. The standard disclosure illustrates an 8,100 won stock being disposed of around a 5,700 won limit-down price. That does not mean the shares are guaranteed to sell 30% lower — the order is priced at the limit to prioritise execution, and the actual fill is set by the order book that day. On a session that gaps down, though, the odds of filling near the bottom rise accordingly.
Collateral top-ups are also restricted: the standard disclosure accepts cash or eligible substitute securities only. Planning to sell holdings to cover the call can fall out of step with the settlement cycle, so choosing the instrument in advance is a different exercise from finding one after the notice arrives.
Rate cards are shaped differently at every broker
Both brokers split the rate into a base and a spread. Mirae Asset shows a 2.89% base plus a 3.01-6.61% spread; Kyobo shows a 2.91% base plus a 1.49-6.49% variable component. The base tracks the broker's own funding cost while the spread is set by the holding tier. Roughly a third of the headline rate is therefore linked to market rates, which means the rate card is a number that gets revised, not a constant.
The widest gaps sit at the top of the table. For 31-60 days the two are effectively identical at 8.5% and 8.60%, but the "up to 7 days" tier splits 4.4% against 5.90% — 1.5 percentage points apart. For money borrowed briefly and repaid quickly, the first two rows of the rate card matter more than the average.
What to watch
- Which interest convention your broker uses — retroactive or tiered. Under the retroactive method, decide on repayment before crossing a boundary day (7, 15, 30, 60, 90)
- The first two rows of the rate card — short-dated borrowing is where brokers diverge most
- Leverage ratio (loan divided by position size) — above 60%, a 16% drop reaches the 140% maintenance line
- The 130% level — breaking it cuts the time to forced liquidation from two business days to one
- Acceptable collateral — cash and eligible substitute securities only; sale proceeds may arrive too late
- Base-rate revisions — about a third of the headline rate moves with funding costs