Korea's Monetary Policy, Part 1: The Toolkit
Eight times a year, seven people vote on one number. The Bank of Korea’s base rate anchors the cost of overnight interbank borrowing, and from there it spreads into mortgage rates, corporate bond yields, and the exchange rate. This post covers how the decision is made, which instruments the Bank of Korea uses to make it stick, and what happens when the usual toolkit runs out. Throughout, I draw on the Bank of Korea’s Monetary Policy in Korea (2017 edition), the institution’s own account of its framework and operations.
Table of Contents
- 0. Eight Times a Year
- 1. The Decision Room
- 2. Three Tools, One Goal
- 3. The Word Is Also a Tool
- 4. When the Floor Disappears
- References
0. Eight Times a Year
The Bank of Korea’s Monetary Policy Board (금융통화위원회, MPB) meets eight times a year to decide the base rate. The calendar is set in advance: January, February, April, May, July, August, October, and November. In the other four months (March, June, September, December) the Board still meets, mainly to review financial stability, but does not make a rate decision. Unscheduled meetings are possible; the BOK held one in March 2020 and cut the rate by 50 basis points.
The Fed’s FOMC also meets eight times a year, but the two processes differ in most other ways.
Figure 1. BOK and FOMC meeting schedules across the year. The BOK holds rate-setting meetings (MP) in eight months and financial stability reviews (FS) in the other four. The FOMC row shows the 2025 calendar (no meetings in February, April, August, or November); the 2026 calendar swaps May for April. Both hold eight rate decisions a year.
FOMC meetings run over two days, and the decision is announced on the afternoon of the second day, U.S. Eastern time. MPB rate meetings are single-day sessions: the decision is announced in the morning, followed by the governor’s press conference. The FOMC releases minutes three weeks after each meeting; the BOK releases its minutes two weeks after, on a Tuesday.
Both calendars matter. Korean markets follow FOMC decisions closely because U.S. rates move capital flows in and out of Korea, the dollar-won exchange rate, and the BOK’s own room to act. A Fed hike that strengthens the dollar puts depreciation pressure on the won, which can import inflation, and the BOK may then feel pressure to tighten even when domestic conditions don’t clearly call for it.
The base rate is a target, not a rate anyone trades at directly. The Board decides where overnight interbank borrowing should price, and the instruments in the next section are what push market rates there. Deciding the rate and implementing it are two separate steps.
1. The Decision Room
The Monetary Policy Board has seven members: the governor (who chairs it), the senior deputy governor, and five members recommended by different bodies (the Minister of Economy and Finance, the BOK governor, the chair of the Financial Services Commission, the chair of the Korea Chamber of Commerce and Industry, and the chair of the Korea Federation of Banks) and appointed for four-year terms. Under the Bank of Korea Act, a decision needs at least five members present and a majority of those present. Because the recommendations come from several bodies, no single one of them controls a majority.
The week before a rate decision follows a routine the public mostly doesn’t see. Staff from the BOK’s main departments brief Board members during the week, and on the day before the meeting a trends briefing (동향보고회의) gives all members an integrated review of domestic and global economic and financial conditions. The rate meeting starts at 9 AM. After discussion and a vote, the governor holds a press conference to explain the decision. Any dissenting members are named when the decision is announced, and the minutes, published two weeks later, summarize the views and arguments raised in the discussion (without attaching names to each view).
If you’re trying to read the BOK’s reaction function, the minutes are often more useful than the decision. By the morning of the meeting, the market has usually priced the outcome. The minutes show how opinion was spread across the Board, how close the call was, and what data would have to change for the majority to move. A 5-2 hold says something very different about the next meeting than a unanimous hold.
The BOK frames its governance around three principles: independence from political instruction, accountability for its decisions, and transparency about its reasoning. They support each other. Transparency creates accountability by making weak reasoning harder to hide, and accountability protects independence by making it costly to look politically responsive. An institution that explains itself publicly in detail is harder to pressure quietly.
How does BOK independence fit with the governor being a presidential appointee? Appointment and instruction are different things. The governor serves a fixed four-year term (renewable once) and can’t be removed over policy disagreements. The government does have two formal channels, though. The Vice Minister of Economy and Finance (or a vice-chair of the FSC) may attend Board meetings and speak, but not vote (열석발언권, Article 91 of the Bank of Korea Act). And under Article 92, if the Minister of Economy and Finance judges that a Board decision conflicts with the government’s economic policy, the Minister can request reconsideration (재의요구). If the Board then re-adopts the same decision with at least five votes, the President makes the final call. This power has been used very rarely, but it is part of the legal framework, and in periods of fiscal stress the line between coordinating with the government and being directed by it depends partly on institutional culture.
2. Three Tools, One Goal
When the Board sets the base rate, it is setting a target. What the BOK actually steers is the overnight call rate, the rate at which banks lend reserves to each other for one day. Three instruments do this work: open market operations, standing facilities, and reserve requirements. In practice, open market operations do most of it.
Open Market Operations
The BOK’s primary instrument is the repurchase agreement (환매조건부채권, RP). An RP is a collateralized short-term transaction: the BOK purchases securities from a financial institution with an agreement to sell them back at a specified price on a specified date. The cash transfer in the initial leg raises the institution’s reserve balance at the BOK, expanding reserve supply in the interbank market and pushing the overnight rate toward the target. RP sales do the reverse.
Figure 2. RP purchase (매입) injects reserves into the banking system by exchanging cash for collateral; RP sale (매각) absorbs reserves by selling securities with an agreement to buy them back. Both transactions unwind at maturity. Because Korea’s banking system has long had surplus liquidity, the BOK’s regular operation has been the 7-day RP sale.
The detail that matters for Korea is the direction. For years, the banking system has had a structural liquidity surplus, so the BOK’s regular weekly operation has been the 7-day RP sale, which absorbs reserves and keeps the call rate from falling below the base rate. RP purchases were used mainly when markets were short of liquidity, for example in 2008 and 2020. That is now changing. As the surplus shrinks (a smaller current account surplus, heavy overseas investment by residents, and FX swaps with the National Pension Service), the BOK announced in 2025 that it would start regular RP purchases to supply liquidity on an ongoing basis.
To absorb the longer-lasting part of the surplus, the BOK issues Monetary Stabilization Bonds (통화안정증권, MSB) with maturities of up to two years. RPs unwind on their own when the agreement matures; MSBs keep the liquidity locked up until the bonds mature or the BOK offsets them. RPs handle short-term swings, and MSBs handle the underlying level.
In a crisis, the scope of open market operations widens. Central banks around the world have accepted a wider range of collateral (from government bonds to covered bonds, mortgage-backed securities, and corporate paper) and lent more against it. If the haircut on a bond is 30%, a bank holding ₩100 billion of it can borrow ₩70 billion; at a 10% haircut, the same bond supports ₩90 billion. The base rate hasn’t changed, but banks can get more central bank liquidity from the same assets, which helps the system absorb funding stress.
Standing Facilities
The second instrument is the standing facility: two windows that are always open, through which banks can borrow from or deposit with the BOK at rates fixed relative to the base rate. Banks can borrow overnight through the lending facility (자금조정대출) at the base rate plus 100 basis points and deposit excess reserves in the deposit facility (자금조정예금) at the base rate minus 100 basis points. Together these form a corridor around the policy target. (The BOK temporarily narrowed the lending side during the 2020 market stress.)
Figure 3. The interest rate corridor formed by the BOK’s standing facilities. The lending facility (base rate +100 bp) sets a ceiling; the deposit facility (base rate −100 bp) sets a floor. By comparison, the ECB’s corridor was historically ±75 to ±100 bp around its main refinancing rate; since September 2024 its deposit rate sits just 15 bp below the main rate and its lending facility 25 bp above.
The corridor holds through arbitrage. If the overnight rate rose above the lending facility rate, banks would borrow from the BOK instead of from each other, which takes pressure off the interbank market and pulls rates back down. If it fell below the deposit facility rate, banks would park reserves at the BOK instead of lending them, draining funds from the market until rates recovered. The BOK doesn’t need to trade every day to keep the rate inside the band.
Banks rarely use the standing facilities. Borrowing at the lending window signals to other banks that you needed emergency central bank money, which can create exactly the reputation problem that made you borrow. Because of this stigma, the facilities work mostly as a credible backstop: they shape rate expectations even on days when nobody uses them.
Reserve Requirements
The reserve requirement system mandates that deposit-taking institutions hold a minimum fraction of their deposits as reserves at the BOK. Required ratios vary by deposit type: 7% on demand deposits, 2% on time and savings deposits, and 1–7% on foreign currency deposits.
The textbook mechanism is the money multiplier. If banks must hold a fraction $r$ of deposits as reserves, each unit of base money supports up to $1/r$ units of deposits. A 10% ratio implies a multiplier of 10; raising it to 20% cuts it to 5. In theory, changing the requirement directly changes how much credit the banking system can create.
In practice, reserve requirements are rarely used as an active policy tool. In the U.S. and the euro area, banks have held reserves far above the legal minimum since 2008, so changing the requirement does little. Korea is different: the BOK absorbs surplus liquidity through MSBs and RPs, so banks don’t sit on large excess reserves. Even so, the BOK hasn’t changed the ratios often in recent years and prefers to steer the call rate directly. The requirement mainly works as a structural buffer, giving banks a stable demand for reserves and supporting payment settlement.
3. The Word Is Also a Tool
The base rate decision, RP operations, and standing facilities all work on the price and availability of money today. A fourth instrument works on something else: expectations about the price and availability of money in the future.
Forward guidance, announcing the expected future path of policy rates, became a standard tool after 2008, when policy rates hit their lower bound, though the Fed and the Bank of Japan had used versions of it earlier. A credible commitment to hold rates at a level for a stated period moves long-term rates immediately, without changing today’s overnight rate. A five-year government bond yield is roughly the expected average overnight rate over five years plus a term premium. If the central bank can credibly shift the expected path of short rates, it moves medium- and long-term yields without any trade in the RP or bond markets.
Forward guidance comes in two forms. Calendar-based guidance names a time horizon: “the base rate will stay at its current level for at least eighteen months.” State-contingent guidance names a condition: “the base rate will stay at its current level until unemployment falls below 6 percent.” Calendar-based guidance is easy to communicate but ties the central bank to a fixed timeline whatever happens. State-contingent guidance is more flexible and says clearly what would trigger a change, but the public has to understand the threshold and trust the central bank to honor it.
Several economists have questioned whether forward guidance works as cleanly as the theory says. Mishkin (2004) argued against central banks publishing their expected rate path, in part because the public tends to read a conditional forecast as an unconditional promise, and Goodhart (2001) discussed why policy committees were reluctant to commit to a future rate path at all. A published path is a forecast that depends on the current outlook. If conditions change and the central bank changes course, which may be entirely sensible, the public can see it as a broken promise rather than appropriate flexibility.
Issing (2005) raised a different concern: credibility. If a central bank publishes explicit guidance and keeps deviating from it, even for good reasons, markets eventually stop treating it as informative. At that point the tool stops working, and the central bank may be worse off than if it had said nothing, because markets now expect explicit guidance that no longer carries information.
Are these critiques against forward guidance in general, or against particular ways of doing it? Mostly the latter. The point isn’t that signaling future intentions is harmful. It’s that the tool is fragile: guidance that is too specific, badly timed, or issued by an institution without a good track record of explaining itself can damage the credibility it was supposed to build.
4. When the Floor Disappears
Everything above assumes one thing: there is room to cut the base rate. RP operations, the corridor, and forward guidance all assume the policy rate can still move down when conditions get worse.
The 2007–2008 global financial crisis removed that room for most major central banks. The Federal Reserve cut its target range to 0–0.25% in December 2008. The ECB cut its deposit rate below zero in June 2014, and its main refinancing rate reached 0% in March 2016. The Bank of Japan had been close to zero since the late 1990s. With conventional cuts used up, central banks needed other tools.
Figure 4. Non-conventional monetary policy instruments used after the 2008 global financial crisis. The BOK used expanded liquidity support in both 2008 and 2020, including large RP purchases and some outright bond purchases in 2020, but did not adopt formal quantitative easing, negative rates, or yield curve control.
Liquidity support to financial institutions and credit markets widens the terms of ordinary central bank lending: more eligible collateral, longer maturities, smaller haircuts. It extends the OMO framework rather than replacing it. The BOK expanded its repo operations in both 2008 and 2020. In April–June 2020 it ran unlimited (full-allotment) RP purchases, which the Korean press called “Korean-style QE”, and it also bought government bonds outright.
Quantitative easing means large-scale outright purchases of government bonds and other securities, expanding the central bank’s balance sheet with no near-term plan to reverse it. It differs from RPs in direction and duration. An RP is short-term and unwinds on its own; QE purchases are long-term and aimed at compressing the term premium on longer-dated bonds. Buying 10-year government bonds takes duration out of the market and adds reserves, pushing long yields down even when the overnight rate can’t fall further. The Fed, ECB, and Bank of Japan ran large QE programs. The BOK’s 2020 bond purchases were small and temporary by comparison, and it never announced a QE program.
Forward guidance as a commitment device looks different at the zero lower bound. A central bank that can’t cut further can still ease financial conditions by credibly promising to keep rates at zero longer than markets expect, which lowers the expected path of short rates and long-term yields today. This is stronger than ordinary rate communication: it is a self-imposed limit on future policy, not just a forecast.
Negative interest rates push the policy rate below zero. The ECB introduced a negative deposit facility rate in June 2014, and the Bank of Japan followed in January 2016. The idea is to charge banks for holding excess reserves so they lend the money instead. Whether this reaches retail depositors depends on whether banks pass the cost on to household accounts, and most haven’t, given the political and reputational risk, so negative rates have mostly squeezed bank margins. The Bank of Japan ended its negative rate in March 2024, and the ECB’s deposit rate returned to positive territory in 2022.
Yield curve control sets an explicit target for the yield at a specific maturity and commits to buying as many bonds as needed to defend it. The best-known example is the Bank of Japan’s target of around 0% for the 10-year government bond yield, introduced in September 2016 and ended in March 2024. QE specifies a quantity; YCC specifies a price, and the central bank stands ready to buy whatever volume holds the yield there.
Through both 2008 and 2020, the BOK was more cautious than the Fed or the ECB. Korea’s policy rate bottomed at 0.5% in 2020–2021, well above zero, and the BOK never adopted negative rates, formal QE, or yield curve control. That reflected both preference and constraint. For a small open economy with significant external financing needs, aggressive balance sheet expansion risks weakening the currency and triggering capital flow swings that are harder to manage than in the U.S. or the euro area. Caution and vulnerability pointed the same way.
Part 2 covers what happens after the decision: the six channels through which the base rate reaches the economy, the lags that force monetary policy to look ahead, and the structural challenges (a flatter Phillips curve, demographic pressure on the neutral rate, and spillovers from global monetary conditions) that have made the toolkit harder to use.
References
- Bank of Korea (2017). Monetary Policy in Korea. Seoul: Bank of Korea. bok.or.kr
- Mishkin, F. S. (2004). Can Central Bank Transparency Go Too Far? NBER Working Paper No. 10829. nber.org/papers/w10829
- Goodhart, C. A. E. (2001). Monetary Transmission Lags and the Formulation of the Policy Decision on Interest Rates. Federal Reserve Bank of St. Louis Review, 83(4), 165–186.
- Issing, O. (2005). Communication, Transparency, Accountability: Monetary Policy in the Twenty-First Century. Federal Reserve Bank of St. Louis Review, 87(2), 65–83.