How does international capital flow?

Michael Kumhof, Phurichai Rungcharoenkitkul and Andrej Sokol

Understanding gross capital flows is crucial for both macroeconomic and financial stability policy. However, theory is lagging behind empirical work, as much of the literature continues to rely on net capital flow models developed many decades ago. Missing from these models is an explicit tracking of the financial records underlying all goods and asset purchases, namely gross balance sheet positions, which in turn requires modelling the principal medium of exchange, bank deposits. Our new model features gross capital flows and offers a fresh perspective on important policy debates, such as the role of current accounts as indicators of financial fragility, the nature of the global saving glut, Triffin’s current account dilemma, and the synchronisation of gross capital inflows and outflows.

What happens when a Foreign resident wants to exchange some of her Foreign currency held in a Foreign bank into Home currency held in a home bank? She would typically first send a payment instrument to the Home bank, drawing it on her foreign currency account. The home bank would then accept the instrument, convert it into Home currency, and credit it to the Foreign resident’s home currency account. Finally, to settle the transfer, the home bank credits its nostro interbank account with the foreign bank, a new claim in foreign currency (Figure 1, top).

The example shows how a capital inflow from foreign to home must be matched automatically by an equally-sized outflow from home to foreign, with the pair of gross flows leaving the current account unchanged on impact. If banks were to eliminate the resulting currency mismatch from their balance sheets, adjustments of residents’ and banks’ balance sheets could also trigger changes in credit and prices that prompt adjustments in real variables. But these balance sheet changes would again have to satisfy the principles of double-entry bookkeeping. The bottom line is that gross capital flows are conceptually distinct from the current account.

All cross-border financial transactions entail such paired balance-sheet entries, including bond purchases (Figure 1, middle). We refer to these as financial flows. The only type of capital flows that show up in the current account are payment flows, which settle purchases of goods and services (Figure 1, bottom). These, however, are typically an order of magnitude smaller, and less volatile, than financial flows.

Most open-economy macroeconomic models gloss over gross balance sheet positions. They ignore the dynamics of deposit money and other money-like instruments, which are intimately linked to gross financial flows and stocks through well-known accounting rules. The analytical shortcut of focusing only on net capital flows often leads to confusion, one prime example being the conflation of foreign financing and saving.

Figure 1: Examples of financial and net payment flows

We address these issues by developing a two-country model that explicitly tracks domestic and cross-border gross flows between banks and households (Figure 2). We put centre stage bank balance sheet dynamics, which can be almost completely disconnected from the real economy, but also an independent source of business cycle fluctuations. We show how net and gross foreign liabilities can send very different signals for a country’s vulnerability to financial shocks. We also show that net foreign financing (the creation of purchasing power for home residents by foreign banks) and foreign saving (foreign output not consumed by foreign residents) are almost completely different concepts and move in opposite directions in response to many shocks. This has major implications for several policy debates.

Is the current account a good indicator of financial vulnerability?

An economy that runs large current account deficits over many years must of course eventually acquire not only large net but also gross foreign liabilities. But even an economy with small net foreign liabilities can be vulnerable if its gross foreign liabilities are large. That’s because in a financial crisis, creditors do not stop financing current accounts, but debt. Moreover, in response to large financial shocks that require the immediate repayment of a sizeable stock of debt, changes in current account flows cannot make any contribution on impact, and only a small contribution over time.

Our model captures these ideas and can provide quantitative insights. For example, a ‘sudden stop’ shock that requires an immediate large repayment of foreign loans, has no immediate bearing on the current account, but can expose pre-existing financial vulnerabilities stemming from excessive reliance on foreign loans. Especially on impact, repayments using existing deposits, refinancing of previously foreign loans with domestic ones, and other balance sheet adjustments are far more important than adjustments in the current account. The latter do occur, but gradually and merely as a by-product of the recession and real exchange rate changes induced by the shock.

Figure 2: Real and financial flows

Global saving glut, or US credit glut?

The global saving glut hypothesis argues that over-abundant foreign saving has been financing US current account deficits, thus contributing to their widening. Bernanke and the subsequent literature typically equate saving to financing and focus on reasons why the true returns to physical saving in less developed countries may be lower than they appear. However, US households do not finance current account deficits with physical saving provided by foreign households, but rather with purchasing power issued by banks that are more likely to be domestic than foreign.

Consider how US imports can be paid for in the real world: first, by transferring existing domestic or foreign bank balances to foreigners, which involves no new financing. Second, by borrowing from domestic banks and transferring the resulting bank balances to foreign households, which involves domestic but not foreign financing. And finally, by borrowing from foreign banks and transferring the resulting bank balances to foreign households. Only the last option involves foreign financing, but in practice it is the least likely option for most domestic residents.

Gross balance sheet positions are critical to discern current account financing patterns. We show that, unlike current account deficits triggered by credit shocks, deficits caused by foreign saving shocks (as in the global saving glut narrative) are not able to reproduce the highly elastic behavior of US credit observed during the saving glut period.

This has stark policy implications. The global saving glut hypothesis puts the onus of adjustment on current account surplus countries. It encourages them to boost aggregate demand to reduce their “excessive saving”. But from a gross-flow perspective, this approach could in fact exacerbate domestic vulnerabilities, by triggering credit booms in surplus countries (see China more recently or Japan in the 1980s). In that light, the onus of adjustment ought to be on deficit countries, to the extent that their “excessive credit” is the reason for their large deficits.  

Is there a Triffin’s current account dilemma?

One version of Triffin’s dilemma posits that a growing world economy requires an increasing quantity of the global risk-free reserve currency. This is taken to imply that the economy issuing the reserve currency must run persistent current account deficits, eventually becoming increasingly indebted to foreigners, until the currency ceases to be risk-free.

This is flawed in both fact and logic. In fact, the US ran almost uninterrupted current account surpluses until the early 1980s while global dollar reserves grew. And in logic, it treats as equivalent changes in the quantity of physical resource flows and in the quantity of currencies. But the creation of dollars only requires digital bank credit, which is independent of physical trade deficits incurred by households and firms, as can be shown in our model. There is therefore no dilemma.

The synchronisation of gross inflows and outflows: an accounting phenomenon?

Gross capital inflows and outflows of individual economies are highly correlated. From the perspective of net flow models, the two legs of such gross flows must result from two separate decisions. In a typical narrative, foreign investors are the recipients of domestic outflows and “send the capital back”, resulting in a build-up of gross positions. But at the aggregate level, such correlation necessarily follows from the accounting rules for financial flows (see our first two examples in Figure 1). The only possible reasons for a less than perfect correlation are a large role for payment flows (whose matching counterpart is not a capital flow but a goods flow), and measurement error.

Balance-of-payments accounting can also shed light on the predominantly positive correlations between sectoral gross capital flows documented by Hélène Rey and other studies. Take a US foreign direct investment (FDI) into Asia: Asia records an FDI inflow, because a US resident purchased a stake in an Asian firm; but the US must also record the corresponding settlement transaction, for example as an interbank debt inflow. Of course, unlike for aggregate gross capital inflows and outflows, the detailed sectorial correlation patterns are not exclusively determined by these mechanics. Gross capital flows have rightly taken center stage in many open-economy macroeconomics debates. We seek to do justice to four of them by providing a framework that carefully distinguishes between foreign financing and saving, and that highlights the crucial role of banks and money. Many other debates deserve to be re-examined in this new light.

Michael Kumhof works in the Bank’s Research Hub, Phurichai Rungcharoenkitkul works at the Bank for International Settlements and Andrej Sokol works at the European Central Bank.

If you want to get in touch, please email us at or leave a comment below.

Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.

2 thoughts on “How does international capital flow?

  1. Isn’t the 3rd option of foreign financing more common then assumed? Like USA centrically it wouldn’t be. But for nearly every other country that need USD I would assume that they finance in USD using eurodollar transactions.

Leave a Reply