South Korea Reserve Money

KRRESERVEMONEY is a raw stock series recorded as reported without any transformation, the won volume of money issued by the central bank and held either as currency by the public or as reserve deposits by depository institutions. The aggregate is the base on which the broader money stock is built, and unlike narrow and broad money, which sum liabilities of the deposit-taking sector, it sums liabilities of the central-banking sector.

The quantity is a financial-accounts item in the lineage of moneyflows accounting, where issued money is entered simultaneously as a liability of the central-banking sector and as a claim of its holder (Copeland 1949; Copeland 1952). Its components sit in the sectoral balance sheets built to measure financial-asset and liability stocks (Goldsmith, Lipsey, and Mendelson 1963), and the same magnitude is read from the financial account of the integrated social-accounting system (Stone 1966). The conventions for sectoring institutional units and classifying instruments follow the handbook treatment (Dawson 1996).

Summing components as heterogeneous as circulating currency and reserve balances into a single won total is an aggregation problem, and aggregation theory combines components weighted by their user costs into an exact index (Barnett 1980; Diewert 1976). The total recorded here, however, is the compiled unweighted sum, and the log-change index tradition merely illustrates the alternative way such components could be combined (Divisia 1925; Fisher 1922).

Conceptually both components are claims that bear no yield or only a low one, so the return forgone in holding them is the price equating present and future income (Fisher 1930). Movements in the aggregate belong to the same family as the money and credit totals studied in the long-run macrofinancial record (Schularick and Taylor 2012). Because the aggregate is a sectoral stock rather than a flow, it is interpreted through the stock-flow accounting identity (Obstfeld and Rogoff 1995), and the series is seasonally adjusted at source and recorded here as an average balance without further adjustment.

KRED applies no transformation to the series, neither deflating, re-scaling, smoothing, nor annualizing it, so each published number is the compiled stock exactly as measured, in billions of won. The seasonal adjustment is performed at source and reproduced verbatim rather than applied by KRED, and the average-of-period timing convention likewise follows the compilation practice at source.

The measurement basis is the flow-of-funds framework, first founded as moneyflows accounting (Copeland 1949) and later implemented as a full empirical system (Copeland 1952), which records issued currency and reserve deposits simultaneously as a liability of the issuing sector and a claim of the holder under quadruple-entry bookkeeping. The aggregate is read from the financial account of the integrated social-accounting system (Stone 1966), its balance-sheet tabulation follows the year-by-year national balance sheets (Goldsmith, Lipsey, and Mendelson 1963), and the practitioner conventions for sectoring and instrument classification follow the handbook treatment (Dawson 1996).

Combining components that differ in liquidity and yield into one won figure invokes aggregation theory, under which quantities weighted by their user costs combine into an exact index (Barnett 1980; Diewert 1976). The index-number apparatus for such weighting is the classical one (Fisher 1922; Törnqvist 1936; Divisia 1925), yet no weighting whatsoever is imposed on this series. The recorded value is an average-of-period balance, and in every case the series exists as an accounting-consistent stock rather than a modelled quantity (Fisher 1930), belonging to the family of national balance-sheet measures (Lane and Milesi-Ferretti 2001).

Reserve money is the aggregate on which the supply decisions of the central bank are inscribed most directly, so it shows at closest range the imprint that policy operations leave on the financial system. Money is a claim whose opportunity cost is the interest forgone by holding it, which links the aggregate to the intertemporal choices of its holders (Fisher 1930). Reading the components as a liquidity-weighted quantity rather than a simple sum brings out more sharply the monetary services the aggregate delivers (Barnett 1980; Diewert 1976).

The moneyflows accounting situates issued currency and reserve deposits within the circular flow of payments (Copeland 1949; Copeland 1952), and the integrated accounts make the holders' balances a direct object of demand analysis (Stone 1966). Whether the gap between this aggregate and the broader totals widens or narrows reveals the extent to which the deposit-taking sector expands central bank money, and such expansions overlap with the episodes that preceded credit booms and financial instability in the long-run record (Schularick and Taylor 2012; Jordà, Schularick, and Taylor 2017).

As a stock the series is one of the sectoral balances measuring the reallocation of purchasing power over time, a reallocation formalized through the saving-investment identity (Obstfeld and Rogoff 1995). Gross sectoral positions carry information beyond what net flows reveal (Obstfeld 2012), the deepening of monetary and financial claims relative to activity is itself a structural indicator (Goldsmith, Lipsey, and Mendelson 1963), and system-wide liquidity conditions the path of the subsequent cycle (Lane and Milesi-Ferretti 2018).

For market participants reserve money is a liquidity overlay for reading funding conditions in the money market, capturing the quantity of central bank money that ultimately satisfies settlement and reserve demand (Copeland 1952). The return forgone on reserve deposits anchors the cash-flow sensitivity of depository institutions to policy-rate moves (Fisher 1930), and whether a balance sits in circulating currency or in reserve deposits carries different meaning under a liquidity-weighted view (Barnett 1980; Diewert 1976).

The pace at which this aggregate expands into the broader totals moves with the phases of credit expansion, and the long-run record linking rapid broad-money growth to crisis frequency gives desks a reference for gauging the probability of disorderly adjustment and for calibrating tail scenarios in liquidity-sensitive portfolios (Schularick and Taylor 2012; Jordà, Schularick, and Taylor 2017).

Liquidity stocks must be read with attention to valuation and measurement, cautions raised by the position-measurement and return-measurement literature (Gourinchas and Rey 2007a; Curcuru, Dvorak, and Warnock 2008), and this caveat generalizes to all gross positions (Obstfeld 2012). The sectoral balance sheets and the gross-position estimation frame the exposures that lenders, money-market investors, and macroprudential overseers monitor when sizing funding risk (Goldsmith, Lipsey, and Mendelson 1963; Lane and Milesi-Ferretti 2007).

KRRESERVEMONEY is a seasonally-adjusted-at-source central bank money aggregate, and its published log-level is the statistical object, expected to be integrated of order one. The sample is 272 monthly observations spanning 2003-10 to 2026-05, and the unit-root triplet is fitted on the log-level with a constant and trend.

The augmented Dickey-Fuller regression of Dickey and Fuller (1979), in the lag-augmented form of Said and Dickey (1984), fails to reject the unit root at p = 0.9363, the nonparametric test of Phillips and Perron (1988) concurs at p = 0.7462, and the stationarity-null test rejects trend stationarity at p < 0.01 (Kwiatkowski et al. 1992), so all three procedures agree on a clean I(1) classification. The GLS-detrended escalation of Elliott, Rothenberg, and Stock (1996) is reserved for ambiguous outcomes under the house protocol and is not invoked on this clean reading.

Because the log-level is I(1), the portmanteau and the break search are run on its first difference, since a level portmanteau reads the near-unit autocorrelation of a stochastic trend and a level mean-break search spuriously segments it (Perron 1989; Hamilton 2018; Bai and Perron 1998). The portmanteau statistic of Ljung and Box (1978), refining the form of Box and Pierce (1970), rejects the white-noise null on the differenced series, with Q = 47.32 at lag 12 and Q = 64.64 at lag 24, both at p = 0.000, and the multiple-break procedure computed by the dynamic-programming algorithm of Bai and Perron (2003) finds no break in the mean of the differenced series.

The seasonal battery is run with the interpretation that the seasonal reading reflects both an inherent property and the source adjustment. The seasonal-unit-root test of Hylleberg et al. (1990), with the monthly mechanics of Beaulieu and Miron (1993), rejects unit roots at the seasonal frequencies with a statistic of 351.08 at p = 0.000, so no stochastic seasonality is present, yet the seasonal dummies are jointly significant with F = 2.94 at p = 0.0011 and the seasonal portmanteau also rejects on the differenced series with a statistic of 95.79 at p = 0.000, so a deterministic seasonal component survives the source adjustment. Since the test outcome differs from the seasonally-adjusted designation carried at source, no clean non-seasonal verdict is assigned, and this reading is consistent with the diagnostic frame of Canova and Hansen (1995) and the seasonal-adjustment literature (Ghysels and Osborn 2001).