Hoarding of international reserves: It's a neighbourly day in Asia

• Author: Xingwang Qian,Eli Remolona,Yin‐Wong Cheung
• Published date: 01 May 2019
• Date: 01 May 2019
• DOI: http://doi.org/10.1111/1468-0106.12297

ORIGINAL MANUSCRIPT

Hoarding of international reserves: It’sa

neighbourly day in Asia

Yin-Wong Cheung

1

| Xingwang Qian

2

| Eli Remolona

3

1

Department of Economics and Finance,

City University of Hong Kong, Kowloon

Tong, Hong Kong

2

Department of Economics and Finance,

SUNY Buffalo State, Buffalo, NY, USA

3

Center for Development Economics,

Williams College, Williamstown,

MA, USA

Correspondence

Yin-Wong Cheung, Department of

Economics and Finance, City University of

Hong Kong, Kowloon Tong, Hong Kong.

Xingwang Qian, Department of Economics

and Finance, SUNY Buffalo State, Buffalo,

NY, 14222 USA.

Eli Remolona, Center for Development

Economics, Williams College,

Williamstown, MA 01267, USA.

Email: yicheung@cityu.edu.hk;

qianx@buffalostate.edu;

eliremolona@gmail.com

Funding information

Research Grants Council of the Hong Kong

Special Administrative Region, China,

Grant/Award Number: CityU 11500617

Abstract

To explain why Asian countries seem to have been hoard-

ing international reserves, especially since the 1997 crisis,

we consider various regional neighbourhood effects. One

such effect is that of “catching up with the Joneses”.We

revisit that effect by analysing several refinements of

it. We also consider the fear of the kind of contagion that

the crisis-hit countries saw in 1997. Finally, we look at the

possibility of a regional financial cycle, in which the con-

ditions that led to the crisis might have been correlated

across countries. We find that refining the Joneses effect

to take account of trade links strengthens its power to

explain the build-up of reserves. We also observe that a

country that finds itself more vulnerable than its regional

neighbours would tend to accumulate more reserves.

Finally, we find that a common regional factor related to

current-account balances spurs further reserve accumula-

tion. Contrary to previous analyses, our results suggest

that only a couple of Asian countries have been holding

excessive reserves. Some were actually holding less

reserves than would be optimal in the presence of

neighbourhood effects.

1|INTRODUCTION

The Asian financial crisis in 1997 was a momentous event for the region. In the five crisis-hit coun-

tries of Indonesia, Malaysia, the Philippines, Thailand and Korea, the crisis brought to a halt a decade

of remarkable growth (e.g. Moreno, Pasadilla, & Remolona, 1998). The crisis ensued in July 1997

Received: 2 December 2018 Accepted: 24 February 2019

DOI: 10.1111/1468-0106.12297

208 © 2019 John Wiley & Sons Australia, Ltd wileyonlinelibrary.com/journal/paer Pac Econ Rev. 2019;24:208–240.

and the next few months saw capital outflows, currency depreciation and stock market collapse in

the crisis countries. Within the year, these countries all found themselves in deep recessions.

One lesson the five countries drew from the crisis was the insurance benefits of adequate reserves.

On the eve of the crisis in June 1997, the five held varying levels of international reserves, ranging

from 5.9% of GDP for Korea to 24.6% of GDP for Malaysia. In the face of capital outflows, these

reserves quickly proved inadequate. Within 6 months, Korea had lost 33% of its reserves, the Philip-

pines 23% and Indonesia 18%. Once they had recovered from the crisis, the five countries embarked

on an unprecedented accumulation of reserves. In June 1997, the five countries together held reserves

amounting to 10.7% of GDP. By December 2006, they had jacked this ratio up to 24.2%. Other coun-

tries seem to have drawn the same lesson. The world’s total international reserves increased steadily

from US$ 1.6 trillion in 1997 to US$ 10.9 trillion in 2015.

Even though both theory and experience suggest that international reserves can reduce the proba-

bility of a financial crisis and, thus, mitigate output loss, they can be costly to hoard (Rodrick, 2006).

Hence, there is an optimal level of reserves beyond which the insurance gains no longer justify the

costs. However, there is little consensus on what this optimal level is, and determining it remains a

challenging task. In much of the literature, the amounts of reserves held by Asian countries are seen

as excessive.

The demand for international reserves can be driven by precautionary and non-precautionary

motives. The precautionary motive would typically be about being able to cover import financing

and external debt payments in the face of changing levels of international reserves (Frenkel, 1974;

Frenkel & Jovanovic, 1981).

1

The Greenspan–Guidotti rule, for example, says that developing econ-

omies should hold sufficient international reserves to cover a year of short-term external debt pay-

ments (Greenspan, 1999).

2

The precautionary motive has been extended to include the role of

reserves as a self-insurance policy to avoid sudden stops, financial instability and crisis-induced out-

put losses.

3

In Krugman’s (1979) model of balance-of-payments crises, the money supply plays an

important role, in which episodes of capital flight lead investors to exchange their money holdings

for dollar reserves held by the central bank. Jeanne and Ranciere (2011) have consolidated all these

considerations into a formula for the optimal level of reserves.

In the case of the non-precautionary motive, Dooley, Folkerts-Landau and Garber (2008) argue

that for some East Asian economies, reserve accumulation was driven by a mercantilist development

strategy. In this view, a development strategy that is based on export-led growth would require an

undervalued currency, and this would lead as a by-product to a build-up of reserves. While Delatte

and Fouquau (2012) and Bar-Ilan and Marion (2009) find evidence of the mercantilist motive of

reserve accumulation, Aizenman and Lee (2007) find that it has little significance in explaining the

rise of international reserves in the post-crisis era.

The literature on the precautionary motive explains reserve accumulation as behaviour that

depends only on factors specific to the economy in question. This overlooks a salient feature of the

1997 Asian crisis, which is that it was a region-wide phenomenon. An exception to this is Cheung

and Qian (2009), who revive the “catching up with the Joneses”effect observed by Machlup (1966).

The effect says that a country’s demand for international reserves will depend on how much its

neighbours hold. In the way Cheung and Qian model the “Joneses effect”, an implicit regional rivalry

gives rise to a competitive hoarding mechanism, which boosts international reserves to levels not

explained by traditional precautionary factors. By maintaining higher reserve levels than its neigh-

bours, a country may avoid being the first one in the region to be attacked by currency speculators

and, thus, give itself time to shore up its defences. Empirical support for such a Joneses effect has

CHEUNG ET AL.209

been reported by Aizenman, Cheung, and Ito (2015), Bird and Mandilaras (2010), Cheung and Sen-

gupta (2011) and Pontines and Yongqiang (2011).

The regional nature of the 1997 Asian crisis was evidently not lost to the countries themselves. It

was apparent that something connected the five countries in the crisis. Whatever that something was

recognized to be, it would have affected reserve accumulation behaviour. For one thing, it may have

reinforced the Joneses effect. Another connection among the countries would be the contagion they

all saw during the crisis. Indeed, Glick and Rose (1999) show that this connection has some basis in

fact. The fear of contagion may have led to the build-up in reserves by Asian countries in the period

since the crisis, and not just by the five hit by the crisis. Other countries in the region, namely China,

India, Japan, Singapore and Taiwan, also built up their reserves aggressively. As a group, the

10 economies together accounted for 39% of total global international reserves in 1997 and 59% in

2015. Indeed, the importance of the fear of contagion in Asia is demonstrated by the Chiang Mai Ini-

tiative Multilateralization (CMIM), a $240 billion group effort by the 13 members of ASEAN+3 to

pool their reserves.

4

In the literature, the actions of these countries are often characterized as excessive hoarding

behaviour.

5

However, this literature has not taken into account the possibility that these countries

were behaving in part to guard themselves against contagion from their neighbours. Indeed, it may

also have been the case that a regional financial cycle gave rise to the conditions that led to the crisis.

These conditions would have been correlated across countries in the region. A sense of these correla-

tions could also have been what motivated the countries in the region to build-up their reserves so

aggressively as a group.

In this paper, we consider various neighbourhood effects in an effort to explain why Asian coun-

tries have accumulated so much in international reserves. One such effect is that of “catching up with

the Joneses”. This paper revisits that effect by analysing several refinements of it. These refinements

include a quadratic version of the variable to investigate a possible levelling off of the Joneses effect.

Other refinements include weighting schemes that take account of economic links between neigh-

bouring countries in the region. We also look at the fear of contagion by examining whether mea-

sures of the vulnerability of neighbouring countries affects a given country's accumulation of

reserves. Finally, we investigate the possibility that the conditions that led to the crisis were corre-

lated across countries in a kind of regional financial cycle. We carry out this investigation by

extracting common economic factors and analysing their effects on the accumulation of reserves.

We focus on 10 Asian countries. These include the five that were hit by the Asian crisis and five

others that have been identified in the literature as having hoarded reserves excessively. Figure 1

shows the rise in the levels of reserves of these 10 economies. To facilitate comparison, reserves are

normalized as ratios to GDP. By this measure, Singapore and Taiwan stand out as the two economies

that had been most aggressive in building up their reserves.

6

By 2009, the group as a whole held

reserves amounting to 40% of their combined GDP. This ratio has since levelled off but it was still

37% in 2015. Figure 1 suggests that the ratio of international reserves to GDP has not been growing

without bound; even the bound can be economy-specific. Indeed, with the decreasing (net) marginal

benefit of holding international reserves, the incentive to follow the Joneses is likely to be

diminishing beyond a certain level. To assess this behaviour, we consider a quadratic Joneses effect

variable. The diminishing marginal effect implies a concave Joneses effect.

We uncover interesting results. We find a significant quadratic Joneses effect, indicating a level-

ling off of that effect. Moreover, we find that a Joneses variable that is constructed with weights

based on trade links outperforms the unweighted version. In considering the fear of contagion, we

find that when an economy finds its current-account position to be weak relative to the position of

210 CHEUNG ET AL.