Korea Overnight Financing Repo Rate

KRKOFR is the raw series recorded without transformation, the daily Korea Overnight Financing Repo Rate computed as a volume-weighted average of interest rates on overnight secured repurchase-agreement transactions collateralized by government and public bonds. A secured overnight rate of this kind is the repo or collateralized-financing rate whose level falls below the general riskless rate when collateral trades special (Duffie 1996), and it occupies the same overnight maturity point that an unsecured interbank rate occupies (Hamilton 1996; Furfine 1999).

Conceptually, this rate is the shortest point of the term structure and the instantaneous riskless rate. The forward-rate decomposition and liquidity premium of longer yields are built upon this point (Hicks 1939), and the empirical measurement of yields across maturities also rests on this anchor (Macaulay 1938).

The discount function and zero-coupon curve are recovered from coupon-bond prices conditional on a near-riskless overnight anchor (McCulloch 1971, 1975). Therefore the parsimonious level-slope-curvature parameterization and the extended forward-rate form of the curve treat this overnight rate as their short-end limit (Nelson and Siegel 1987; Svensson 1994; Gürkaynak, Sack, and Wright 2007).

Operationally, the rate is the operating target that monetary-policy liquidity operations steer within a standing-facility corridor (Bindseil 2004; Borio 1997; Bartolini, Bertola, and Prati 2002).

KRED applies no transformation, so the published value is the secured overnight repo rate exactly as recorded, with no smoothing, deflation, or rescaling. The measurement basis by which such a rate comes to exist is the transaction-weighted aggregation of executed overnight repurchase-agreement rates, the secured-financing measurement in which the rate is the collateralized counterpart of the riskless overnight benchmark and falls below general collateral when bonds trade special (Duffie 1996).

The aggregation follows the transaction-level identification and volume-weighting logic developed for the overnight interbank rate (Furfine 1999; Hamilton 1996). Each constituent is quoted on the simple money-market yield-to-maturity convention that maps a price to an annualized return (Macaulay 1938).

Because the overnight rate is the operating target, its recorded path reflects daily liquidity-management operations conducted within a standing-facility corridor (Bindseil 2004; Borio 1997; Bartolini, Bertola, and Prati 2002).

The same recorded overnight rate is the short-end input to the discount-function and zero-coupon term-structure estimators that recover yields and forward rates from coupon-bond prices. These estimators are implemented as cubic or exponential splines (McCulloch 1971, 1975; Vasicek and Fong 1982), as the parsimonious or extended parametric forms (Nelson and Siegel 1987; Svensson 1994), or as the daily off-the-run smoothing now standard in curve construction (Gürkaynak, Sack, and Wright 2007). Bliss (1997) catalogues the competing estimators, and the zero-coupon term structure so obtained feeds duration measurement (Fisher and Weil 1971).

As the overnight riskless rate, KRKOFR is the operating target through which monetary policy is transmitted, the rate that liquidity operations hold near the policy setting inside a corridor of standing facilities (Bindseil 2004; Borio 1997). Its daily determination reflects reserve-supply and demand conditions in the overnight market (Hamilton 1996; Bartolini, Bertola, and Prati 2002), and its microstructure is read from transaction-level activity (Furfine 1999).

Because it is the shortest maturity, the rate anchors the expectations-plus-liquidity-premium decomposition that links it to longer yields (Hicks 1939). It further supplies the short-end limit of the measured yield curve (Macaulay 1938; Nelson and Siegel 1987; Svensson 1994; Gürkaynak, Sack, and Wright 2007).

Deviations of the secured rate from general-collateral levels signal collateral scarcity and repo specialness (Duffie 1996). This is information about funding conditions that the discount-function curve estimated from coupon prices propagates to every maturity (McCulloch 1971).

In valuation, KRKOFR is the secured overnight discounting and floating-rate reference benchmark, so it underlies the discount factors applied to fixed-income cash flows and the floating leg of overnight-indexed swaps. The repo rate it measures is the actual cost of collateralized financing of bond positions, where specialness lowers the rate on scarce collateral (Duffie 1996). The same overnight rate determines the carry on leveraged positions whose interest-rate exposure is summarized by duration (Macaulay 1938; Fisher and Weil 1971).

Discounting and immunization both require the full zero-coupon term structure built outward from this overnight anchor (McCulloch 1971, 1975; Vasicek and Fong 1982). This term structure is expressed through the parametric and forward-rate curves (Nelson and Siegel 1987; Svensson 1994; Gürkaynak, Sack, and Wright 2007), whose construction methods are compared by Bliss (1997).

Hedging and relative-value desks read its level against the policy corridor that pins it (Bindseil 2004), and read its microstructure against the parallel unsecured overnight rate (Hamilton 1996; Furfine 1999).

On the 1,113 daily observations spanning 2021-11-25 to 2026-06-09, fit with a constant and trend, the unit-root battery agrees that the Korea Overnight Financing Repo rate is integrated of order one. The augmented Dickey-Fuller test of Dickey and Fuller (1979), in the ARMA-consistent lag-augmented form of Said and Dickey (1984) and with lag length set as in Ng and Perron (2001), does not reject a unit root with p = 0.4389, the nonparametric Phillips and Perron (1988) test concurs with p = 0.687, and the Kwiatkowski et al. (1992) stationarity test rejects its trend-stationary null at p < 0.01, so the verdict is an unambiguous I(1). The GLS-detrended power escalation of Elliott, Rothenberg, and Stock (1996) is reserved for ambiguous outcomes under the house protocol and is not required on this clean reading.

Because the level is integrated, the mean-shift and serial-correlation diagnostics are run on the first difference, the stationary object those procedures require, since a break search or a portmanteau on an integrated level would spuriously segment and read near-unit autocorrelations (Bai and Perron 1998; Perron 1989). The multiple-break procedure of Bai and Perron (1998), computed by the dynamic-programming algorithm of Bai and Perron (2003), finds no break in the mean of the differenced series, consistent with the parameter-instability inference of Andrews (1993). The Ljung and Box (1978) portmanteau statistic, refining the original Box and Pierce (1970) form, is computed on the first difference and rejects the white-noise null at lags 10 and 20, with Q = 82.67 and Q = 125.33 and p = 0.000 and p = 0.000, and the automatic portmanteau test of Escanciano and Lobato (2009) concurs with a statistic of 12.67 at p = 0.000.

The series is daily and has no low-integer seasonal period, so the seasonal-unit-root machinery of Hylleberg et al. (1990) and the Canova and Hansen (1995) seasonal-stationarity test are inapplicable and are deliberately not run, the degeneracy of the seasonal auxiliary regression at a daily period being the standard ground (Beaulieu and Miron 1992; Ghysels and Osborn 2001).