Korea's Stage-3 Stress DSR: What 1.5%p Cuts From Your Limit

Take out a mortgage in the Seoul metropolitan area and the bank assesses your repayment capacity at your actual rate plus 1.50 percentage points. That is stage 3 of the stress DSR regime, in force since July 1, 2025. You do not pay more interest. You are simply allowed to borrow less.

Plug in an annual income of 60 million won, a 30-year equal-payment mortgage and an actual rate of 4.0%, and the limit falls from 418.9 million won to 352.2 million won. That is 66.7 million won, or 15.9%, gone. What follows is where those figures come from, and how they move when income and rates change.

How the stress rate is set

The stress rate is calculated as the gap between the highest lending rate of the past five years and the current rate, subject to a floor of 1.5% and a ceiling of 3.0%. In recent years the calculation has sat on the floor, which is why the Financial Services Commission's "base stress rate of 1.5%" functions in practice as a fixed number.

A stage multiplier is then applied: 25% at stage 1, 50% at stage 2, 100% at stage 3. The 1.50 percentage points loaded onto metro-area mortgages is 1.5% times 100%. A loan-type multiplier can apply on top. Pure floating-rate loans take the full amount, while hybrid and periodic-reset products take less the longer the rate is fixed — which is why a five-year periodic product yields a larger limit than a pure floater. If you go floating, it is worth understanding how the new-origination and outstanding-balance COFIX benchmarks diverge.

Coverage widened at stage 3. Mortgages, credit loans and other lending are all in scope at both banks and non-bank institutions, with credit loans caught only where the outstanding balance exceeds 100 million won.

How much of the limit disappears

A DSR of 40% means annual principal and interest cannot exceed 40% of annual income. Inverting the present-value formula for an equal-payment loan gives the limit.

limit = monthly capacity × (1 − (1+i)^−n) ÷ i (i = monthly rate, n = 360 months)

Assuming an actual rate of 4.0%, a 30-year term and no existing debt, varying only income. Because the stress rate applies solely to the limit calculation, the monthly payment actually due stays at the 4.0% level.

It is often said that higher earners lose more. The table shows that holds only for the amount. The reduction rate is 15.9% on every row. Income enters the formula linearly through monthly capacity while the rate enters through the annuity coefficient. Stress DSR does not discriminate by income band — it compresses everyone's limit by the same proportion.

For reference, the FSC estimated at the time of stage 2 that limits would fall by 3–9% for mortgages at banks and non-banks and 1–2% for bank credit loans. The 8.5% reduction implied by the regional stage-2 column above sits at the top of that range, because the assumptions here — 30-year floating with no existing debt — maximize the effect.

The stress rate is not interest you actually pay. What you owe stays the same; only what you can borrow shrinks.

Lower rates get cut harder

One result runs against intuition. The same 1.50 percentage points bites differently depending on the underlying rate — and it bites harder when rates are low.

Limits per one million won of monthly repayment capacity, by rate:

At 3.5% the cut is 16.4%; at 5.0% it is 15.1%. The spread is modest, but it reveals the regime's character. Stress DSR grips tighter in an easing cycle. When the policy rate falls and lending rates follow, limits should expand — and the regulation claws back a fixed share of that expansion. Falling rates and rising borrowing capacity do not necessarily arrive together.

Why regional mortgages are still at 0.75%p

Stage 3 was not applied uniformly. Mortgages on property outside the capital region still carry the stage-2 load of 0.75 percentage points. Herald Business reports that the carve-out, originally due to expire in June 2026, has been extended again to the end of December. Greened News notes this is the third extension, after end-2025 and mid-2026.

The rationale is a divergence in market conditions: prices and transaction volumes are recovering in the capital region but not elsewhere. For a homebuyer, the operative detail is that the distinction turns on where the collateral property sits, not where the borrower lives. On the 60 million won income row, the gap between the metro and regional limits is 31.2 million won.

The carve-out has an expiry date. Once stage 3 reaches regional mortgages after December 31, 2026, the same income will support a smaller loan again. The way policy loan products and regulated lending compute limits differently is easier to see alongside how Didimdol and Bogeumjari loans split on income and ceiling.

What to check

• Location of the collateral property — 1.50%p in the capital region, 0.75%p in non-regulated regional areas. It follows the property, not your address
• Rate type — pure floating takes the full stress load. Hybrid and periodic-reset products take a lower share the longer the fixed period, yielding a larger limit at the same income
• Existing credit loan balance — above 100 million won it picks up the stress rate too, eating further into mortgage capacity. Work out whether it can be brought under 100 million before the mortgage is drawn
• December 31, 2026 — the expiry of the regional carve-out. If you are looking outside the capital region, limits differ either side of that date
• Each bank's actual figure — the calculations here assume a 30-year equal-payment loan with no existing debt. Term, repayment structure, current obligations and the applicable DSR ceiling (40% at banks, 50% at non-banks) all change the answer, so run the numbers at two or more lenders

Sources

• Stage-3 stress DSR takes effect in July — Korea.kr policy briefing
• Second-half stress DSR operating direction — Financial Services Commission
• Regional mortgages get another stage-3 reprieve to year-end — Herald Business
• Regulators grant regional mortgages a further six-month delay — Greened News