Current Accounts in Debtor and Creditor Countries

The paper asks “what is the current account response to a transitory income shock such as a temporary improvement in the terms of trade, a transfer from abroad or unusually high production?” (Introduction). Its answer departs sharply from the textbook intertemporal prediction: “the model generates the novel prediction that favourable income shocks lead to current account deficits in debtor countries and current account surpluses in creditor countries” – rather than the conventional prediction of a surplus everywhere, since standard models without capital accumulation only allow a saved income shock to be invested in foreign loans.

Helpman and Razin aside, the authors’ own thought experiment: if a country saves a favorable transitory shock and allocates the marginal unit of wealth among domestic capital and foreign loans in the same proportions as the average unit of wealth, then “since by definition the share of a debtor country’s wealth invested in domestic capital exceeds one, an increase in wealth (savings) results in a greater increase in domestic capital (investment), leading to a deficit on the current account.” The mirror-image logic applies to creditor countries, where the wealth increase exceeds the increase in domestic investment, with the remainder invested abroad, producing a surplus. The paper is explicit that this sharp result rests on three assumptions: the shock is saved (consumption smoothing), foreign capital investment is unavailable to the country (removable without changing the qualitative result), and the marginal unit of wealth is allocated like the average unit – the bulk of the paper’s theoretical effort goes toward assessing that third assumption.

The model features infinitely many atomistic countries with Poisson-switching “good” and “bad” productivity regimes (π̄ vs. π̲) and continuously stochastic Brownian output shocks within a regime (Eq. 1-2); investors have log utility (Eq. 3) and choose a share of wealth to hold as risky domestic equity versus riskless foreign loans (Eqs. 5-6). Market clearing for international loans pins down the world interest rate as r = π̄ − σ² (Eq. 7), where π̄ is average world productivity, and the resulting capital-stock rule (Eq. 8) shows the country’s capital stock rising in both its productivity and its wealth – crucially, investment risk (parameterized by σ) prevents the “usual corner solution of a world of deterministic linear economies in which all the capital is located in the country or countries that have the highest productivity.” Since high-productivity countries borrow to exploit their better investment opportunities while bearing the resulting risk, “high-productivity countries are debtors…while low-productivity countries are creditors” (Eq. 10), and the volume of borrowing is larger the larger are cross-country productivity differentials and the lower is investment risk.

Because investors hold logarithmic utility, the equilibrium investment rule implies an arbitrage condition π_j = r + σ²·(k_j/a_j) (Eq. 14): the expected return to domestic equity must equal the world interest rate plus a risk premium equal to the covariance between the return on equity and the return on the investor’s total wealth. When a positive output shock raises wealth without changing the profitability of investment, the only way the marginal unit of new savings can satisfy this unchanged arbitrage condition is if it is invested in exactly the same proportions as the existing portfolio – which is precisely the “marginal equals average” assumption from the introduction’s thought experiment, now derived rather than assumed. The paper notes a testable corollary: the model predicts the equity premium σ² + π̄ − π should be systematically larger in debtor countries, a prediction it flags as “new” and untested.

Taking the stochastic process for net foreign assets (Eq. 11) and its expectation (Eq. 12), the paper decomposes the expected current account into a consumption-tilting term and a mean-reversion term, and shows that positive output shocks (dθ_j > 0) lead to current-account deficits in debtor countries and surpluses in creditor countries, “since a + (π_j − π̄)/σ > 0.” Productivity shocks (dπ_j) operate through two separate channels: an income effect formally identical to the output-shock effect, and a much larger rate-of-return effect, since a positive productivity shock in a creditor country raises the expected return to domestic equity and triggers a portfolio reallocation (an investment boom) rather than merely an income effect. Because “rate-of-return effects…consist of reallocations in the stocks of assets, their effects on the current account are much larger (of order 1) than the income effects of the same shocks (of order dt)” – so infrequent productivity shocks should show up as large spikes in current-account time series, while frequent output shocks drive its ordinary period-to-period variance (Eq. 13).

Generalizing to Stone-Geary preferences (risk aversion varying with wealth via a parameter β_j) and riskless labor income (λ_j), the paper derives a modified investment rule (Eqs. 17-18) and shows “a simple rule which determines when a positive output shock leads to a current account deficit: the country’s debt has to exceed a threshold that depends on (1) how attitudes towards risk vary with wealth, and (2) the size of labour income.” This threshold, −(λ_j+β_j)/(π̄−σ²), can be positive or negative and collapses to exactly zero in the constant-relative-risk-aversion, no-labor-income benchmark. Intuitively, if relative risk aversion decreases with wealth (β_j<0), a positive shock induces riskier investment positions and “the share of the marginal unit of wealth invested in risky domestic capital exceeds its share in average wealth,” potentially turning some creditor countries into deficit-runners after a favorable shock; conversely, if labor income is riskless and sizable (λ_j>0), a wealth increase raises the ratio of financial to human wealth, inducing a more conservative reallocation that can produce current-account surpluses even in some debtor countries.

Using IMF international-investment-position data for thirteen OECD countries (1971-1993), the authors construct an “adjusted debt” measure (net external debt minus outward FDI and foreign equity holdings) and find that “in six out of seven countries where adjusted debt is positive, the current account is countercyclical…while in five out of six countries where adjusted debt is negative, the current account is procyclical” – a simple cross-country correlation between debt and current-account cyclicality of −0.54 (Table 2, Figure 1). Splitting the pooled country-year sample at various debt thresholds and testing for equality of slopes, “current accounts are clearly procyclical in creditor countries and countercyclical in debtors,” with the difference “significant at the 5-10 percent level” across a wide range of thresholds (Figure 2a), a pattern that strengthens in the 1982-1993 subperiod and survives dropping the years immediately following the 1973 and 1979 oil shocks.

The authors consider two alternative stories consistent with correlated international business cycles: that OECD expansions raise world interest rates and so strain debtor countries’ service obligations, and that revaluation of existing cross-border capital holdings alone could generate the pattern given debtor countries’ smaller gross foreign asset positions. Controlling for growth in 6-month LIBOR and OECD-wide (excluding own-country) per capita GNP growth, the cross-country correlation between adjusted debt and current-account cyclicality “remains negative, and is equal to -0.55,” though the range of debt thresholds at which the debtor/creditor cyclicality difference is statistically significant narrows once these global shocks are partialled out (Table 3, Figure 4) – the authors report this as a weakening but not elimination of the core pattern.

The authors state plainly: “Sovereign risk and foreign investment are two important features of real economies that we have left unmodelled here,” though they “feel confident…that our results would survive all but the most extreme versions of models that incorporate these elements.” More fundamentally, “the most glaring omission of this paper is the absence of adjustment costs to investment”: because investment is “bang-bang” in the model, “the effects of all these forces are collapsed into a single instant, effectively precluding any meaningful discussion of the dynamic aspects of the current account responses to shocks,” and the authors note they are extending the theory to incorporate sluggish investment responses as follow-up work.

This summary is grounded in the paper’s own MIT Department of Economics working paper draft (No. 97-12, July 1997), the direct precursor to the November 2000 Quarterly Journal of Economics publication; the working paper’s title, authors, abstract, and argument match the published article (confirmed via the QJE’s own listing, vol. 115(4), pp. 1137-1166), so the two are treated here as the same paper under normal pre-publication revision over the roughly three-year gap between the two.