January 08, 2008
International Reserves: Messages from the ASSA
I did not get a chance to go to too many sessions at the ASSA meetings in New Orleans (the AEA agenda is here). That being said, I did manage to squeeze in a few on international economics, and the topic of several papers was foreign exchange reserves.
In the AEA session "International Reserves", my former colleague Joshua Aizenman and coauthor Jaewoo Lee examined in Financial versus Monetary Mercantilism-Long-run View of Large International Reserves Hoarding [pdf] the various hypotheses concerning the accumulation of foreign exchange reserves. Noting that the conjecture that East Asian countries were accumulating reserves essentially for "self insurance" was not plausible after 2003, they noted the debate over mercantilism versus hoarding as motivations.
An interesting hypothesis is that in the presence of competitive hoarding between symmetric emerging markets (in an attempt to maximize exports to a target market), the effects of the hoarding are dissipated, so that there is in some sense excessive hoarding.
They then observed the oddity that mercantilism is not persuasive as an explanation for the recent accumulation of reserves since the East Asian economies could have long been characterized as mercantilist. They reconciled this inconsistency by observing that in the earlier period, financial mercantilism (reducing the cost of capital through the banking sector) might have been prevalent, while in the later period, it was through monetary mercantilism, i.e., a weak exchange rate.
Just in case I made mistakes in taking notes, here's the abstract from the paper:
The sizable hoarding of international reserves by several East Asian countries has been frequently attributed to a modern version of monetary mercantilism -- hoarding international reserves in order to improve competitiveness. From a long-run perspective, manufacturing exporters in East Asia adopted "financial" mercantilism-subsidizing the cost of capital -- during decades of high growth. They switched to hoarding large international reserves when growth faltered, making it harder to disentangle the monetary mercantilism from precautionary response to the heritage of past financial mercantilism. Monetary mercantilism also lowers the cost of hoarding, but may be associated with negative externalities leading to competitive hoarding. From this viewpoint, this paper makes three observations on the East Asian reserve accumulation. First, the recent large hoarding of reserves in Japan and Korea occurred in the aftermath of the growth strategy that combined export promotion and credit subsidization (financial mercantilism). Second, whether the ultimate motive is mercantilist or precautionary, the ongoing reserve hoarding in Asia contains an element of competitive hoarding, which is likely to have negative externalities among countries involved. Finally, China’s hoarding of reserves partly reflects the precaution against the financial fragility that is likely to follow the slowing of economic growth.
The discussant, Tom Willett observed that the extent of reserve accumulation seemed to be too great to be explained by mercantilism, and attributed the phenomenon to the attempt to avoid economic adjustment that might be accompanied by political instability.
In their paper entitled "The Optimal Level of International Reserves for Emerging Market Countries?", Olivier Jeanne and Romain Ranciere forwarded a theoretical utility-based model of reserve holdings. Thus armed they are able to conduct welfare analysis, and thus identify, in principle, the optimal level of reserves.
The motivation for holding reserves in their model is an external credit constraint on borrowing imposed on a small open economy. Foreign exchange reserves constitute an insurance contract; if there is no sudden stop, there is a negative transfer to the country. If there is a sudden stop, then there's a positive one.
The derivation of the solution indicates that optimal reserves are a function of ratio of short term debt to GDP, and the cost to GDP of a sudden stop.
(This paper is not online, as far as I know. A related paper by Olivier Jeanne is published in BPEA here (sub.req.)
The discussant Linda Goldberg lauded the paper as a clearly useful analysis. But she highlighted several issues. The first is that foreign exchange reserves were nowhere linked to fixed exchange rates (except insofar as a fixed exchange rate regime could increase the probability of a "sudden stop" in capital flows), the typical motivation for reserve holdings.
Eduardo Levy-Yeyati presented a paper entitled "Cost of Reserves", that seeks to rationalize the accumulation of reserves. He notes that reserves not only reduce the probability of a crisis, but also thespred paid on the full stock of sovereign debt, reducing the marginal cost of reserve accumulation.
The magnitude of the benefit is estimated using a 2 factor model of EMBI spreads, as show in Gonzalez Rozada and Levy-Yeyati, EJ 2007 (working paper version here [pdf]).
From the conclusion:
While these findings do not deny the fact that self-insurance is costly and should be considered as a second best solution in a context of imperfect international financial markets, they certainly shed a different light on the cost-benefit that should inform the decision about the optimal amount of reserves. Needless to say, these estimates could be refined to take into account country specific characteristics (the currency and maturity composition of sovereign debt, among other things, should certainly influence the impact of liquid reserves). However, they help to illustrate the main message of this note, namely, that the cost of reserves, as typically measured, may have been considerably overstated.
There was one more paper in the session agenda I wasn't able to see, entitled "Allocation of International Reserves Across Asset Classes" by Richard Portes, Elias Papaioannou, and Grigorios Siourounis. This paper is not online, but one with a similar title is here.
[Joshua Aizenman tells me this last paper was cancelled. Instead, the paper was "Fear and Market Failure: Global Imbalances and 'Self-insurance'" by Marcus Miller and Lei Zhang. Here is the discussant's comments, by Joshua Aizenman.]
In different session, my former colleague Yin-Wong Cheung presented a paper entitled "Hoarding of International Reserves: Mrs Machlup's Wardrobe and the Joneses" [pdf]. From the introduction:
We postulate that the international reserve accumulation process pertaining to the Mrs Machlup's Wardrobe metaphor may serve some relevant economic purposes. It is quite noncontroversial to state that, on the other things being equal basis, international reserves help absorb unexpected (external) shocks and smooth current and capital account imbalances. The crisis experience and the development after the crisis appear to be consistent with the notion of accumulating international reserves to forestall future speculative attacks. The question, of course, is how high the level of international reserves an economy has to hold?
On his wife's dress need, Machlup (1966, p. 26) suggests that it depends "on the Joneses with whom she wishes to keep up." That is, besides some fundamental considerations, the buildup of international reserves depends on the behavior of neighboring economies. Ignoring the question of why Mrs Machlup has to keep up with the Joneses for a moment, the (implicit) rivalry among economies may give raise to competitive hoarding mechanism that pushes the holding of international reserves to a level that is difficult to be explained by only traditional considerations.
Besides the pure psychological desire to feel good and not to be perceived as inferior, there are a few reasons why economies would like to keep up with their peers. Remarks by Feldstein (1999) and Fischer (1999), for example, offer some insight on the keeping up with the Joneses motivation. After the crisis, these two noted economists observed that economies with a higher level of international reserves survived the East Asian financial crisis better than those with a lower level. Thus, a level of international reserves that is relatively higher than your neighbors may diffuse the speculative pressure on your own economy and divert it to the neighboring economies and, hence, reduce the chance of bearing the full cost of an attack. In other words, when a financial crisis is brewing in the region, if two economies have similar economic fundamentals, the one with a higher level of international reserves is less likely to be attacked and more likely to survive the crisis. Another reason for keeping up with the Joneses is that international reserves can have a positive impact on an economy’s output prospects. If the level of international reserves is a barometer of financial heath, an economy has an additional incentive to keeping up with the Joneses to position itself to compete for international capital and foreign direct investment, which tend to have a level of productivity proficiency higher than the domestic capital. For developing economies, the output effect of international reserves also arises from their ability to reduce costs of borrowing in the international capital market and provide needed liquidity when there is a reversal of capital flows. A relatively high level of international reserves will, thus, provide a catalyst for economic growth and enhance output prospects, which in turn will improve the market sentiment and, hence, reduce an economy's vulnerability to attack.
There are other international economics papers I saw presented, dealing with the optimal exchange rate regime, exchange rate pass through, and economic integration in East Asia. More on that in a future post.
Posted by Menzie Chinn at January 8, 2008 08:52 AMdigg this | reddit
There's a an endogeneity problem with the "keeping up with the Joneses" paper. Aren't burgeoning reserves themselves a reflection of higher FDI?
If the level of international reserves is a barometer of financial heath, an economy has an additional incentive to keeping up with the Joneses to position itself to compete for international capital and foreign direct investment, which tend to have a level of productivity proficiency higher than the domestic capital.
Posted by: Emmanuel at January 8, 2008 02:06 PM
One problem with the "lowers borrowing cost argument" -- namely, emerging economies in aggregate aren't borrowing right now. I don't quite see how you can argue brazil gains by sterilizing in real at over 10% to buy Treasuries at 4%! the net gain isn't obvious.
Do Aizenman and Lee flesh out how Chinese fx reserves help protect against the kind of financial vulnerabilities China currently has, as opposed to the Asian crisis style problem of too much short-term external debt or Argentina's problem with domestic liability dollarization. I see how reserves could help to manage a run out of the banking system into foreign assets -- though with a 10% of GDP current account surplus, china could absorb a huge run pretty easily without dipping into reserves.
on the flip side, I see a lot of evidence that excessive reserve accumulation is starting to damange the banking system, not the least because China's government is now forcing the banks to hold $ as part of their reserves in order to slow reserve growth and reduce the costs of sterilization. And i think that is generally true - lots of countries have reverted to hiking mandatory reserves (a tax on the banks) to help with sterilization. Not to mention the side effects of negative real rates ...
Posted by: bsetser at January 8, 2008 06:03 PM
It seems clear that some Asian economies are following the old Japanese model - use an undervalued currency to spur economic growth. My guess is that other explanations of their FX reserves fall far short in explaining the current levels. What is less clear is what will happen when growth in aggregate demand in the importing counties slows and they start objecting to the practice.
Posted by: drecon at January 9, 2008 07:17 PM