Journal Issue

Seven Questions on the Role of Precautionary Savings in Open Economies

International Monetary Fund. Research Dept.
Published Date:
March 2011
  • ShareShare
Show Summary Details

Are precautionary savings important to understanding open-economy issues? This article reviews recent literature on the relevance of precautionary savings for global imbalances and more generally for the size and allocation of capital flows.

Question 1: Why should we consider precautionary savings in open economies?

The early literature on open economies has largely ignored precautionary savings, defined as the amount of wealth accumulated to self-insure against stochastic income fluctuations instead of to smooth deterministic income changes. Prior to the adoption of computational techniques in economics, researchers relied extensively on the assumption of perfect foresight to obtain analytical solutions. Even more recent contributions using numerical methods often use first-order linear approximations that eliminate precautionary motives. However, Ghosh and Ostry (1994, 1997) show that precautionary savings can significantly influence the current account, the understanding of which is of paramount importance in the study of open economies. Being equal to the difference between saving and investment, the current account should indeed improve during periods of higher volatility that stimulate the accumulation of precautionary savings. Supporting empirical evidence is found using data for the Unites States, Japan, and the United Kingdom.

Question 2: Did precautionary savings play a role in the surge of global imbalances?

Recent studies emphasize the importance of precautionary savings for understanding the growth of global imbalances. Fogli and Perri (2006) develop a two-country, business-cycle model to quantify the optimal reduction in US net foreign assets in response to the “Great Moderation.” Their results suggest that the decline in business-cycle volatility experienced by the United States between 1984 and 2005 accounted for 20 percent of US external deficits.

Precautionary savings are likely to have played an even stronger role in the accumulation of assets by emerging countries. Aizenman and Lee (2007) find empirical evidence for the hypothesis that the self-insurance motive has been an important determinant for the hoarding of international reserves. Durdu, Mendoza, and Terrones (2009) provide further support using a calibrated and small open-economy model. They find that self-insurance motives against sudden stops can indeed lead to the accumulation of large precautionary reserves.

Finally, precautionary savings have also stimulated the accumulation of international assets by commodity-exporting countries. Bems and de Carvalho Filho (2009) use a calibrated model to show that oil and gas exporters should accumulate a considerable amount of foreign assets not only to deal with the exhaustibility of their resources, but also to self-insure against large fluctuations in commodity prices.

Question 3: What is the optimal amount of precautionary savings?

The growing interest in precautionary motives has raised questions about the optimal level of net foreign assets (NFA) that countries should hold to self-insure against shocks. Jeanne and Sandri (forthcoming) address this issue by considering the dynamic stochastic optimization problem of a small, open economy subject to GDP fluctuations. To guarantee the existence of an optimal finite amount of precautionary savings, a country needs to be “impatient” so that in the absence of uncertainty, it would prefer to frontload consumption by borrowing internationally. Otherwise, precautionary motives, by providing an incentive to accumulate savings, would lead to an infinite stock of foreign assets. The optimal NFA position is then given by the level at which the precautionary motive to save exactly counterbalances the impatience desire to borrow.

Unfortunately, it is hard to provide a robust estimate for this level, since the degree of impatience crucially hinges on the relation between the unobservable intertemporal discount factor and the interest rate that is set exogenously in a small, open economy model. Sandri (forthcoming) addresses these issues by solving for a world general equilibrium in which the interest rate is pinned down endogenously and the discount factors are common across countries. This leads to fairly robust predictions for the optimal NFA positions of six regions, countries or country categories in the world (the United States, euro area, Japan, China, oil exporters, and the rest of the world) that are assumed to differ only with respect to GDP volatility calibrated on 1980-2008 data. The observed heterogeneity in volatility creates a wide dispersion in optimal NFA positions, with the most volatile oil-exporting countries accumulating international assets up to several multiples of GDP.

Question 4: How important is it to reach the optimal amount of precautionary savings?

Jeanne and Sandri (2011) show that there is an important upside to the difficulty in pinning down a robust value for the optimal NFA. The extreme sensitivity of NFA to the degree of impatience implies that welfare-wise is not crucial to achieve the exact optimal level. Calibration exercises show that countries can rather effectively self-insure against income volatility by simply targeting whichever NFA positions they currently have, the key issue being to avoid drifting toward excessive borrowing. This insight is also confirmed in the general equilibrium analysis in Sandri (forthcoming), which reveals that while the heterogeneity in volatility across countries can lead to a wide dispersion in optimal NFA positions, it can account for only very small current account imbalances. This is because countries can already achieve substantial self-insurance with their current NFA positions and move only very gradually toward their optimal levels. An important exception regards countries close to their international borrowing constraints: given that negative shocks cannot be easily smoothed out by issuing additional debt, it becomes essential to run large current account surpluses to quickly accumulate precautionary savings. This is especially relevant for those countries that also face a possible tightening of the borrowing constraint.

Question 5: How large are the welfare gains from hedging instead of relying on precautionary savings?

Self-insurance through the accumulation of precautionary savings comes at the cost of postponing consumption. For some shocks, countries can more efficiently seek insurance by using hedging instruments. Borensztein, Jeanne, and Sandri (2009) consider the welfare gains that commodity exporters could seize by hedging against price fluctuations rather than accumulating precautionary savings. Hedging conceptually provides two distinct sources of benefits. First, it leads to more stable consumption by curbing export income volatility. Similar to the literature on the welfare costs of the business cycle, welfare gains from this channel are moderate. Second, it reduces the need to hold precautionary savings and allows the country to finance a temporary increase in consumption by borrowing internationally against more secure future export revenues. This can lead to large welfare gains, equivalent to a permanent increase in consumption of several percentage points.

Question 6: Is within-country idiosyncratic risk relevant for international precautionary savings?

Precautionary savings in an open economy are often modeled as driven by self-insurance motives against aggregate shocks. However, given the limited extent of within-country risk sharing, idiosyncratic volatility is also relevant, especially because it is much larger than aggregate volatility. For example, Sandri (2010) shows that high idiosyncratic entrepreneurial risk can explain why—contrary to benchmark neoclassical growth models—growth accelerations in developing countries often lead to an improvement in the current account. This is because to self-insure against idiosyncratic investment risk, entrepreneurs accumulate precautionary savings that can sustain large and persistent capital outflows. On the contrary, if domestic financial markets are able to allow for sufficiently high risk sharing (e.g., through well-functioning equity markets), entrepreneurs need less precautionary savings and growth accelerations can actually involve large capital inflows.

Question 7: Can precautionary motives explain international portfolios?

Precautionary motives can be useful to explain not only countries’ NFA positions, but also their portfolio composition. Mendoza, Quadrini, and Ríos Rull (2009) show that countries with lower domestic financial development tend to invest foreign savings in precautionary safe assets, while countries with better financial markets seek riskier foreign investment opportunities. As also discussed in Gourinchas, Rey, and Govillot (2010), this has important implications for the return performance of countries’ international portfolios. Better domestic risk sharing or lower risk aversion can indeed lead to higher returns on assets than on liabilities, as the country invests in foreign high-return risky projects financed by issuing low-return safe liabilities.


    Aizenman, Joshua, and JaewooLee, 2007, “International Reserves: Precautionary vs. Mercantilist Views, Theory and Evidence,”IMF Working Paper 05/198 (Washington: International Monetary Fund).

    Bems, Rodolfs, and Irineu deCarvalho Filho,2009, “Current Account and Precautionary Savings for Exporters of Exhaustible Resources,”IMF Working Paper 09/33 (Washington: International Monetary Fund).

    Borensztein, Eduardo, OlivierJeanne, and DamianoSandri, 2009, “Macro-Hedging for Commodity Exporters,”IMF Working Paper 09/229 (Washington: International Monetary Fund).

    Durdu, Ceyhun Bora, EnriqueG. Mendoza, and MarcoTerrones, 2009, “Precautionary Demand for Foreign Assets in Sudden Stop Economies: An Assessment of the New Mercantilism,”IMF Working Paper 07/146 (Washington: International Monetary Fund).

    Fogli, Alessandra, and FabrizioPerri, 2006, “The Great Moderation and the U.S. External Imbalance,”Monetary and Economic Studies, Vol. 24 (December), pp. 20934.

    Ghosh, Atish R., and JonathanD. Ostry, 1994, “Export Instability and the External Balance in Developing Countries,”IMF Staff Papers, Vol. 41 (June), pp. 21435.

    Ghosh, Atish R., and JonathanD. Ostry, 1997, “Macroeconomic Uncertainty, Precautionary Saving and the Current Account,”IMF Working Paper 92/72 (Washington: International Monetary Fund).

    Gourinchas, Pierre-Olivier, HélèneRey, and NicolasGovillot, 2010, “Exorbitant Privilege and Exorbitant Duty,”IMES Discussion Paper 2010-E-20. Tokyo: Institute for Monetary and Economic Studies.

    Jeanne, Olivier, and DamianoSandri, 2011, “Precautionary Savings in the Open Economy: Exact Solution and Approximate Rules” (unpublished; Washington: International Monetary Fund).

    Mendoza, Enrique G., VincenzoQuadrini, and José VíctorRíos Rull, 2009, “Financial Integration, Financial Development, and Global Imbalances,”Journal of Political Economy, Vol. 117 (June), pp. 371416.

    Sandri, Damiano, 2010, “Growth and Capital Flows with Risky Entrepreneurship,”IMF Working Paper 10/37 (Washington: International Monetary Fund).

    Sandri, Damiano, forthcoming, “Precautionary Savings and Global Imbalances in a World General Equilibrium,”IMF Working Paper (Washington: International Monetary Fund).

Other Resources Citing This Publication