South Korea Financial Account Balance

KRBOPFIN is a raw monthly series recorded as reported without any transformation, the net financial-account balance of the resident economy stated in millions of U.S. dollars. The account records the net acquisition of financial assets less the net incurrence of liabilities across direct investment, portfolio investment, financial derivatives, other investment, and reserve assets, so it captures how cross-border transactions are financed rather than what is traded.

The balance is a double-entry accounting magnitude in the lineage of national social accounting, where every cross-border transaction is entered once as a credit and once as an offsetting debit (Meade and Stone 1941). The financial account sits within the integrated income, capital, and financial accounts (Stone 1966), and its definitional and consistency rules descend from the modern social-accounting framework (Stone 1947) and the national-income measurement it formalized (Kuznets 1941), in the sector-by-sector sources-and-uses tradition of moneyflows accounting (Copeland 1949) and the full empirical construction that followed it (Copeland 1952).

Because the financial account is the financing counterpart of the current and capital accounts, it mirrors, up to errors and omissions, the resident economy's net lending to the rest of the world, the intertemporal accounting relation linking saving, investment, and the external balance (Obstfeld and Rogoff 1995). Cumulated financial-account transactions map into the international investment position once valuation change is separated out (Lane and Milesi-Ferretti 2001; Gourinchas and Rey 2007b), and the gross positions behind the net figure carry information the net flow cannot (Obstfeld 2012).

In the underlying accounts every claim is entered as an asset of one sector and a liability of another, and the stock counterparts are tabulated on the national balance sheets built to record financial assets and liabilities (Goldsmith, Lipsey, and Mendelson 1963), under the compilation conventions for sectoring units and classifying instruments (Dawson 1996).

Because the series is a signed flow rather than a stock and is not seasonally adjusted, it retains the calendar structure of cross-border settlement, and it is recorded here exactly as measured without further adjustment.

KRED applies no transformation to KRBOPFIN, neither deflating, re-scaling, smoothing, annualizing, nor seasonally adjusting it, so the published number is the compiled balance exactly as measured, in millions of U.S. dollars. Because the balance takes both positive and negative values, no logarithmic or ratio transform is applied either, and the level is served as recorded.

The measurement basis is the double-entry balance-of-payments framework, in which every cross-border transaction is booked simultaneously as a credit and a debit so that the accounts close (Meade and Stone 1941), and which sits within the integrated system of social accounts (Stone 1966; Stone 1947). The reliability bounds on such estimates were formalized alongside the accounts themselves (Stone, Champernowne, and Meade 1942), and the national-income measurement they extend is the foundational one (Kuznets 1941).

The financial account is compiled instrument by instrument under quadruple-entry bookkeeping, each claim being recorded simultaneously as one sector's asset and another's liability (Copeland 1949; Copeland 1952), following the flow-of-funds handbook treatment for sectoring cross-border transactors and classifying instruments (Dawson 1996), with the stock counterparts tabulated on the national balance sheets (Goldsmith, Lipsey, and Mendelson 1963).

The stock counterpart of the flow is the international investment position, estimated by the external-wealth methods that reconcile flows with positions (Lane and Milesi-Ferretti 2001; Lane and Milesi-Ferretti 2007), with gross positions constructed component by component (Gourinchas and Rey 2007a) and transactions separated from position change through the valuation channel (Gourinchas and Rey 2007b). In every case the published figure exists as an accounting-consistent flow rather than a modelled quantity (Obstfeld and Rogoff 1995).

The financial account is among the most-watched external-sector quantities, because it records the form in which the external balance is financed. The accounting identity that equates the current account with national saving less domestic investment leaves the financing side to the financial account (Obstfeld and Rogoff 1995), and the cumulated transactions map into the net foreign asset position, whose dynamics also run through valuation (Lane and Milesi-Ferretti 2001; Gourinchas and Rey 2007b).

Balances of similar size can differ greatly in composition, so the internal structure of the account carries separate information, since inflows financed by direct investment differ in reversibility from inflows financed by short-term other investment, and the sheer size of gross claims and gross liabilities conditions how capital-flow shocks transmit (Obstfeld 2012; Lane and Milesi-Ferretti 2018).

The underlying financial accounts locate each claim within the circular flow of payments (Copeland 1949; Copeland 1952; Stone 1966), while the national accounts and national balance sheets connect those flows to the measurement of output and income and to the recording of stocks (Kuznets 1941; Goldsmith, Lipsey, and Mendelson 1963).

Rapid accumulation of external claims and liabilities has repeatedly preceded financial stress in the long-run macrofinancial record, so the scale and composition of the financial account inform assessments of external and systemic risk (Schularick and Taylor 2012; Jordà, Schularick, and Taylor 2017). Because the accounts are compiled by double entry, each credit has an offsetting debit, so a current-account surplus appears as an equal net acquisition of foreign assets, the closure property that makes the accounts informative for external-sustainability analysis (Meade and Stone 1941).

For market participants the financial account is a macro overlay for currency and external-funding risk, since the net acquisition and net incurrence of cross-border claims are what ultimately present themselves as demand for and supply of the currency. The form of the financing and its reversibility matter for the exchange rate and for the pricing of country risk, and persistent external deficits raise the probability of a disorderly adjustment (Obstfeld and Rogoff 1995; Obstfeld 2012).

Gross external positions and their measured returns are sensitive to the estimation method, so they must be read alongside the flows with attention to valuation, and the valuation channel through which positions move independently of flows is itself a source of gains and losses on the external balance sheet (Gourinchas and Rey 2007a; Gourinchas and Rey 2007b; Curcuru, Dvorak, and Warnock 2008).

What drives capital in and out is ultimately the intertemporal price of funds and the differential in expected returns (Fisher 1930), so desks read the composition of the financial account for the direction of currency and rates exposure and for funding vulnerability, while external-position estimates bound the size of that exposure (Lane and Milesi-Ferretti 2001; Lane and Milesi-Ferretti 2007; Lane and Milesi-Ferretti 2018).

Because rapid growth in external and credit aggregates has been linked to crisis frequency, the level and momentum of these flows are used to calibrate tail scenarios (Schularick and Taylor 2012; Jordà, Schularick, and Taylor 2017), and the double-entry closure guarantees that a deficit is financed by an offsetting net inflow (Meade and Stone 1941).

KRBOPFIN is the Korean financial-account balance, a signed monthly flow that takes both positive and negative values, so its published level is the statistical object and the log-level transform applied to strictly positive aggregates is not applicable here. The sample is 557 monthly observations from 1980-01 to 2026-05, and the unit-root triplet is fitted on the level with a constant and trend.

The augmented regression of Dickey and Fuller (1979), lag-augmented in the manner of Said and Dickey (1984), fails to reject the unit root with a statistic of −1.444 at p = 0.848, while the nonparametric test of Phillips and Perron (1988) rejects it with a statistic of −15.285 at p = 0.000, and the stationarity-null test of Kwiatkowski et al. (1992) rejects trend stationarity with a statistic of 0.296 at p < 0.01. The augmented regression and the stationarity-null test both point away from clean stationarity while Phillips-Perron dissents alone, a split traceable to the near-unit moving-average structure of the strongly seasonal level that inflates the Phillips-Perron rejection, so under the house honest-agreement protocol, which treats Phillips-Perron as a conservative veto, no clean order-of-integration label is assigned and the reading is reported as ambiguous. The GLS-detrended escalation of Elliott, Rothenberg, and Stock (1996) and the modified M-tests of Ng and Perron (2001) are the designated power and size supplements reserved for such an ambiguous verdict.

Following the matrix specification for signed members, the portmanteau and the break search are computed on the first difference, since a level portmanteau reads the induced autocorrelation of the trending level 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 = 295.36 at lag 12 and Q = 460.139 at lag 24, both at p = 0.000, and the multiple-break procedure computed by the dynamic-programming algorithm of Bai and Perron (2003) locates no break in the mean of the differenced series.

Because the series is not seasonally adjusted, the seasonal battery interrogates genuine calendar structure rather than residual adjustment error. 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 53.354 at p = 0.000, so the seasonal variation is deterministic rather than stochastic, and the seasonal-dummy-F test confirms stable deterministic seasonality, jointly significant with F = 11.160 at p = 0.000. The seasonal-frequency portmanteau returns the lag-24 statistic of 460.139 at p = 0.000 on the differenced series, concentrating the residual autocorrelation at the seasonal lags. This reading of deterministic, stationary seasonal structure is consistent with Canova and Hansen (1995) and with the seasonal-adjustment literature (Ghysels and Osborn 2001).