Invest1 publisher2 min readPublished
Bank of Korea traces 41% of Korea-US long-rate co-movement to American inflation
The Bank of Korea's decomposition of 26 years of daily 10-year yield changes gives US inflation a 41% share of Korea-US rate synchronization, against 18.3% for Fed policy. The transmission runs through policy expectations.
The Investor · Invest desk
What happened
- The Bank of Korea released a report titled "Analysis of Korea-U.S. Long-Term Interest Rate Synchronization" on the 20th, according to en.sedaily.com.
- The study is built on daily changes in 10-year government bond yields in South Korea and the United States from 2000 to 2025.
- US inflation came out as the single biggest factor behind the two countries' long-rate synchronization, contributing about 41% of it.
- Fed monetary policy accounts for 18.3%, behind US long-term interest rates at 22.7% and ahead of US economic conditions at 18%.
- Around major Fed quantitative easing announcement dates, both short- and long-term Korean rates fell, and the long-rate decline came more from policy expectations than from risk compensation.
Compiled by The InvestorSomething wrong?How this is made
Why it matters
- decision A Seoul rate desk ordering its calendar by expected variance gets the US CPI release first and the FOMC statement second, because the study weights one at 2.24 times the other.
- exposure The four published shares sum to 100, which leaves a Korean 10-year position priced, in this model, by four American variables and no domestic one.
- constraint If US shocks reach Korean long yields through expectations of policy rates, the BOK can pull its long end away from Washington only by changing what markets expect it to do with its own rate.
US inflation's 41% share is about 2.24 times the 18.3% the Bank of Korea assigns to Federal Reserve monetary policy [4][6][12]. Add the 22.7% it gives US long-term rates and the two non-policy US variables account for 63.7% of the explained co-movement [5][14].
The study splits transmission two ways. "We analyzed the transmission channels of rate synchronization between Korea and the U.S. by dividing them into a policy channel and a risk-compensation channel, and found that U.S. inflation mainly affected Korean long-term rates through the policy channel," the Bank of Korea said [8]. The policy channel is the path in which an external shock feeds into expectations for a central bank's current and future policy rates and moves the expectations component of long yields [9]. The risk-compensation channel adjusts the risk premium paid to investors, moving the term premium [10].
Read that way, a US price release is traded in Seoul as a forecast of policy rates, and the 18.3% left to Fed monetary policy is nearer to the surprise that survives the data. The Fed still matters in this picture. Its reaction function is the thing the market is forecasting when it reads the CPI.
Sample composition would weaken the CPI-first conclusion. Synchronization was strongest immediately after the 2008 global financial crisis and during the 2021 pandemic inflation shock [3], two episodes in which US inflation was the global shock. The published account gives shares for the whole sample without separate figures by period [16]. Scope is the other problem. The work rests on daily changes in 10-year government bond yields from 2000 to 2025, a 26-year window [2][15]. It explains how the two markets move together, not how much of the Korean yield level is set at home.
On the study's own weights I would rank the US CPI release ahead of the FOMC statement as the higher-variance date for a Korean 10-year book [4][6]. The counter-thesis is testable within a year. If the next run of US yield news comes from supply and fiscal deficits instead of inflation surprises, the 22.7% assigned to US long-term rates should grow at the expense of the 41%, and the ordering reverses [4][5]. The decomposition itself cannot settle that, because it is fitted on a quarter-century in which two inflation episodes carried the synchronization [3].
What to watch
- Whether the BOK publishes the underlying model and subsample contributions, particularly for the years since 2022.