Dealing with Volatile Capital Flows: The Case for Collective Action
If nothing else, the financial crisis of 2007–10 has reinforced the general proposition that unbridled capital flows can pose a potential problem for the global financial system. The seriousness of that problem depends on economic and financial circumstances, including the strength and quality of macroeconomic and prudential institutions and policy tools. Therefore, a respectable case can be made for controls on capital flows.
But the effectiveness of controls on capital inflows also depends on circumstances. Selective controls may be effective, at least temporarily, in influencing the composition of capital inflows, but less effective in limiting their total volume and, therefore, in limiting any affects on exchange rates and other asset prices.
Capital flows, and controls on them, create externalities for the country imposing them, as well as for other countries and the global economic and financial system. Thus, there is sufficient basis for collective action, or actions, in this area. But to invoke the too familiar cliché, the devil is in the details.
One reason for caution is that there is less agreement on three other points:
- Whether substantially more systematic data collection on capital flows is called for.
- In what respects should the scope of surveillance of capital flows be expanded?
- To what extent should the multilateral framework governing capital flows be enhanced? The range of choices is from an IMF ex post blessing or tolerance of the use of capital controls to an implicit advocacy of a return to a world in which essentially all capital flows are tightly controlled at the national level subject to international rules and regulations. (The latter extreme is implicit in some commentaries that are grounded on the view that the benefits of modern finance are limited or negative.)
In my recent remarks at a conference on international monetary policy sponsored by the Swiss National Bank and the International Monetary Fund, I addressed four questions that follow from these points of nonagreement: (1) Why is our focus primarily on external financial flows involving only developing and emerging-market countries? (2) How strong is the case for increased data collection? (3) How much effective policy space do countries have for temporary controls? (4) How strong is the case for revisiting the multilateral framework governing capital flows?
(1) Why is our focus primarily on external financial flows involving only developing and emerging-market countries?
If we had a broadly agreed answer to this question, it would help to inform our understanding of external flows of capital, in contrast with flows from outside the country. I can think of four reasons for the asymmetry. First, the adverse consequences of internal flows are less significant. This is, perhaps, because second, countries generally have more comprehensive and consistent national prudential regimes. Third, countries have a wider range of mechanisms for managing the adverse consequences of flows within the domestic economy and financial system. Fourth, the costs of controls on internal capital flows are higher and their effectiveness is lower.
Turning to the second part of my question and the focus only on developing and emerging-market economies, can external financial flows be a problem for an advanced country? The answer is yes. Consider the ongoing Greek tragedy, of course. Consider Japan over the years, including the more than a decade in which near-zero short-term official Japanese interest rates distorted its own and the world economy and financial system. Consider the United States. You should not be surprised by my mention of the United States! Over the years, many observers have suggested that the United States should impose controls on capital inflows to reduce our chronic current account deficits. Why were those suggestions not taken seriously? The answer lies in issues of stigma, collateral damage, and perhaps dependency on such flows given that the gross stock of foreign claims on the United States economy is more than 150 percent of US GDP.
(2) How strong is the case for increased data collection?
As an economist and former participant in the policy process, I like to have more data sooner. No doubt there is scope for further improvement. However, I am skeptical whether the current availability of data on gross capital movements is a major deficiency. Better international data for macroprudential purposes are a separate, though very complex, matter. In contrast, we now have substantial, comprehensive information on most types of gross stocks and flows of capital. This is a consequence of efforts coordinated under the auspices of the Bank for International Settlements (BIS), in which I participated for many years, and a consequence of the more recent push under the aegis of the IMF to collect and publish information on international investment positions. (This is an area with many entrenched interests among the data collection bureaucracies and those from whom the data are collected. The IMF’s Sixth Balance of Payments and Investment Position Manual took seven years of work.)
With respect to timeliness, I doubt that the increased availability data on the nature and source of cross-border financial flows on an hourly, daily, weekly, or even monthly basis would alter the content and timing of decisions about capital flows. The inherent volatility of the data ensures that policymakers will wait quarters, if not years, before deciding to do something about flows.
It is possible that the recent actions by countries such as Brazil are a counter example, but those actions appeared to be triggered, I would submit, by the behavior of exchange rates—not by the observed behavior of capital flows, though Brazil has an elaborate system of monitoring capital flows.
On the source of flows, how is a policymaker aided by such information? A dollar is a dollar, regardless of where it comes from. Its source does not indicate whether it is good or bad.
Finally, I have long rejected the simple notion of a hierarchy of capital flows in which direct investment is good and banking flows are bad. The data on flows of foreign direct investment, portfolio investment, and bank lending to emerging-market countries in the recent crisis demonstrate that all three types of recorded flows were part of the boom and the bust. Moreover, portfolio and direct investments can trigger problems that are not recorded in the data on gross flows because of associated financing, hedging, and profit-remittance activities.
(3) How much effective policy space do countries have for temporary controls?
I have no fundamental intellectual problem with capital controls as a policy tool. However, I remain skeptical that temporary controls are a powerful tool that can be universally employed.
First, temporary controls, once imposed, have a tendency to become semi-permanent even as their costs increase and their effectiveness declines. This is because controls generate various rents, and those who collect the rents want to continue to do so. This is true even if controls should be time bound as has been suggested by some. (The United States imposed controls on capital flows—mostly outflows—starting in the mid-1970s and it took years to remove them even after the decision had been made to do so.)
Second, I doubt the speed with which temporary controls can be imposed compared with other policies, in particular macroeconomic policies, including exchange rate policies. An exception may be when a country has a permanent structure of controls with the barriers normally set at zero, so that they can be raised if there is a tsunami. However, maintaining such a structure raises many other issues of economic structure and efficiency.
Third, what people would really like are preemptive controls for use before the tsunami hits. But, as is implicit in my skepticism about the usefulness of real-time data, a preemptive regime of controls is a conceptual challenge even to think about.
Fourth, selective controls induce distortions and problems, as Thailand recently discovered. One has, again, the issue of disentangling the good from the bad capital flows. The simplest example is trade credit, which is easy to talk about, in theory, but, in practice, is next to impossible to define and to isolate.
(4) How strong is the case for revisiting the multilateral framework governing capital flows?
To end on a positive note, the case for revisiting the multilateral framework governing international capital flows is strong. In other words, international monetary reform in this area should be explored.
The possibility of amending the IMF Articles of Agreement as they apply to capital flows and capital controls, which were dropped in the late 1990s, should be reconsidered. The process should start with a full-fledged examination, covering much more than the issues we have touched upon at the conference, to see if there is enough common agreement. As an optimist, I think there will be. I envisage an amendment with three components.
First, the amendment should state the long-term goal of complete freedom of capital movements among countries, along with appropriate prudential regulations internally and globally, analogous to the situation within countries. But there should be no timetable explicit or implicit to achieve this goal.
Second, the amendment should guide national policies in terms of both the rights and the responsibilities of IMF members. For example, controls should be permitted but, as much as possible, they should be applied on a nondiscriminatory, national treatment basis.
Third, the amendment should describe and prescribe the role of the IMF management, staff, and members in conducting surveillance over capital flows and controls. The principal rationale for collective action is that both flows and controls generate externalities.
It would be worth considering whether the IMF should be empowered formally to "bless" capital controls (and whether such a power to bless would be ex ante or ex post) or merely to tolerate and provide advice about them. However, the amendment should remove the present anachronistic asymmetry in article VI between current account transactions and capital account transactions and the use and nonuse of IMF resources to finance each type of transaction.
My support of reform in this area is qualified. Reinforcing and updating the multilateral framework governing capital flows should not be separated from other desirable reforms of the international monetary and financial systems. In particular, IMF surveillance on this topic should be explicitly strengthened and extended covering the full range of policies of recipient countries that affect private and official capital inflows and outflows, which will be very intrusive. Importantly, and I suspect even more controversially, IMF surveillance on this topic should be explicitly extended to the full range of policies of source countries that affect capital inflows and outflows, including, for the central bankers, monetary policies.
Adapted from a presentation at an international monetary conference sponsored by the Swiss National Bank and the International Monetary Fund in Zurich on May 11.