Tax on an individual bond attaches to the coupon and nowhere else. Interest income is withheld at 15.4% (14% income tax plus 1.4% local surtax), but the gain from buying below par and being redeemed at par carries no tax at all. Hankyung notes that bond capital gains sit outside Korea's enumerated taxable income, leaving only the interest portion exposed. Two bonds with identical pre-tax yields therefore diverge after tax, and the one with the lower coupon comes out ahead.
Set two bonds side by side — 10 million won face value, three years remaining, both priced to a 3.5% pre-tax yield to maturity — and the after-tax annualized returns land at 3.26% and 2.89%. That is a 0.37 percentage point gap. Tracing the after-tax cash flows of a 1.5% coupon against a 4.0% coupon shows exactly where the gap opens up.

Tax Attaches to the Coupon Rate, Not the Return
Bond returns arrive through two channels: periodic interest set by the coupon rate at issuance, and the difference between purchase price and sale price (or redemption value). Per Toss Bank's summary, interest on an individual bond is withheld at 15.4% while capital gains are exempt.
What matters is that the tax base is the coupon rate applied to face value, not the actual return. Whether you paid 9.4 million or 10.5 million won for a 10-million-won bond carrying a 1.5% coupon, the taxable interest is 150,000 won a year either way. Any excess return earned by buying cheap arrives at redemption as a capital gain — and that portion is never taxed.
This is why government bonds issued during the low-rate years, with coupons locked in low, get discussed as tax-management instruments. Hankyung reported that as of February 2025 the most heavily net-bought issue among individuals was KTB 01500-5003 (20-2), a 30-year bond issued in 2020 at 1.5%, with the 1.125%-coupon KTB (19-06) cited for the same reason. The same article put individual net bond purchases at 5.57 trillion won, 76.1% above the 3.17 trillion won of net stock buying over the period.
One caveat: this structure applies to directly held individual bonds. Bond ETFs are taxed on a holding-period basis, with 15.4% applied to both distributions and capital gains. The same bonds in a different wrapper produce a different tax bill.

Two Bonds at 3.5% Pre-Tax — How Far Apart After Tax?
Hold the conditions constant: 10 million won face value, three years to maturity, annual coupon payments, and a 3.5% pre-tax yield to maturity for both. Under those terms the theoretical price moves with the coupon.
- 1.5% coupon: purchase price 9,439,700 won (94.40% of par) — redeemed at 10 million won, for a 560,300 won redemption gain
- 4.0% coupon: purchase price 10,140,100 won (101.40% of par) — redeemed at 10 million won, for a 140,100 won redemption loss
Strip 15.4% from each coupon payment and stack three years of cash flows:
| Item | 1.5% coupon | 4.0% coupon |
|---|---|---|
| Purchase price (10M won par) | 9,439,700 won | 10,140,100 won |
| Annual interest (pre-tax) | 150,000 won | 400,000 won |
| Annual tax at 15.4% | 23,100 won | 61,600 won |
| Annual interest (after tax) | 126,900 won | 338,400 won |
| Three-year after-tax interest | 380,700 won | 1,015,200 won |
| Redemption gain/loss (untaxed) | +560,300 won | -140,100 won |
| Three-year after-tax profit | 941,000 won | 875,100 won |
| After-tax annualized return | 3.26% | 2.89% |
Both started at an identical 3.5% pre-tax, yet after-tax profit differs by 65,900 won — 0.37 percentage points annualized. The reason is plain arithmetic: over three years the 4.0% bond pays 184,800 won in tax on its coupons while the 1.5% bond pays only 69,300 won. Shifting return into the untaxed redemption gain thickens what survives.
When pre-tax yields match, the tax code favors the lower coupon. The question is whether that premise survives contact with the actual order book.
The after-tax annualized figures come from the internal rate of return on after-tax cash flows. For the 1.5% coupon the flows run -9,439,700 / +126,900 / +126,900 / +10,126,900 won, and the discount rate that zeroes them out is 3.26%. The same method yields 2.89% for the 4.0% bond. Neither assumes coupons are reinvested; adding reinvestment narrows the disadvantage of the higher-coupon bond somewhat. Read alongside the hold-to-maturity comparison for retail-only treasury bonds to see how the tax point shifts across instruments.

The 20 Million Won Threshold, and How Fast the Coupon Burns It
Once annual financial income — interest and dividends — exceeds 20 million won, the excess is aggregated with other income and taxed at progressive rates. As KB explains, tax that would have ended at 15.4% withholding moves into the graduated brackets. What counts toward that threshold is interest actually received, not unrealized bond appreciation. Capital gains, being exempt, never enter the calculation.
Put 100 million won into bonds under the same terms (three years remaining, 3.5% pre-tax YTM) and the annual taxable interest splits sharply by coupon. The figures below account for the fact that a below-par price buys more face value with the same cash.
| Coupon rate | Face value purchased (won) | Annual taxable interest (won) | Annual tax at 15.4% (won) | Share of 20M won threshold used |
|---|---|---|---|---|
| 1.0% | 107,531,000 | 1,075,300 | 165,600 | 5.4% |
| 1.5% | 105,935,000 | 1,589,000 | 244,700 | 7.9% |
| 2.75% | 102,146,000 | 2,809,000 | 432,600 | 14.0% |
| 4.0% | 98,618,000 | 3,944,700 | 607,500 | 19.7% |
Same 100 million won, same pre-tax yield, but the rate at which the aggregation threshold is consumed differs by a factor of 3.6. A holder of the 4.0% bond spends a fifth of the annual allowance on that single position; the 1.0% bond uses 5.4%. For someone already carrying meaningful deposit interest or dividend income, that gap can matter more than the 0.37 percentage points of after-tax yield.
Where This Calculation Breaks Down
Everything above rests on the assumption that pre-tax YTMs are equal. In practice that is the first thing to fail. When tax-driven demand crowds into low-coupon issues, prices rise and their pre-tax yields settle below comparable bonds — you hand back at the price what you saved on the tax. The number to check before buying is not the coupon rate but the pre-tax YTM gap against other issues of the same remaining maturity.
Second is rate direction. A lower coupon means longer duration, so the same rate move swings the price further. If you intend to hold to maturity the interim price is noise, but any chance of selling early turns that sensitivity into realized profit or loss. Running the duration math that converts a rate move into a price move first is the safer order of operations.
Third is the asymmetry of losses. If capital gains are untaxed, capital losses are equally unrecognized — there is no offsetting them against other income the way equity losses can be handled elsewhere. Fourth is the wrapper problem noted above: bond ETFs and bond funds tax capital gains too, so none of this logic carries over.
Finally, for anyone well below the aggregation threshold with modest financial income, the entire benefit collapses to the difference in 15.4% withholding. In that case the thin liquidity and wide bid-ask spreads typical of low-coupon issues can eat whatever the tax code gave back.

What to Watch
- The pre-tax YTM gap — line the candidate up against ordinary issues of the same remaining maturity. A spread of 0.3 percentage points or more may mean the tax advantage has already been priced away
- Your total annual financial income — how much headroom remains under the 20 million won threshold largely determines whether the low-coupon structure pays
- Intended holding period — anything short of hold-to-maturity requires a separate duration sensitivity calculation
- Purchase price versus par — the discount to face value sets the size of the return that flows through untaxed
- Exchange bid-ask spread and volume — if early sale is possible, calculate how much the spread erodes the after-tax edge
- Instrument form — whether it is a direct bond holding or an ETF/fund decides whether capital gains are taxed at all
