"Liquidity Risk Aversion, Debt Maturity, and Current Account Surpluses: A Theory and Evidence from East Asia"

Abstract

After a series of crises, many developing countries came to recognize that reducing liquidity risk is an important self-protection. However, they have alternative strategies for the self-protection. The purpose of this paper is to show that macroeconomic impacts might be very different depending on which strategy developing countries will take. In the first part, we investigate what macroeconomic impacts an increased aversion to liquidity risk can have in a simple open economy model. When the government keeps foreign reserves constant, an increased aversion to liquidity risk reduces liquid debt and increases illiquid debt. However, its macroeconomic impacts are not large, causing only small current account surpluses. In contrast, when the government responds to the shock, the changed aversion increases foreign reserves and may lead to a rise of liquidity debt. In particular, under some reasonable parameter set, it causes large macroeconomic impacts, including significant current account surpluses. In the second part, we provide several empirical supports to the implications. In particular, we explore how foreign debt maturity structures changed in East Asia. We find that many East Asian economies reduced short-term borrowings temporarily after the crisis but increased short-term borrowings in the early 2000s. Since short-term debt is liquid debt, the instantaneous change after the crisis is consistent with the case where only private agents responded to increased aversion to liquidity risk. However, accompanied by substantial rises in foreign exchange reserves, the change in the early 2000s is consistent with the case where the government also started to respond. We discuss that our results have important implications for the recent deterioration in the U.S. current account.

Full text

Page 1

NBER WORKING PAPER SERIES
LIQUIDITY RISK AVERSION, DEBT MATURITY, AND CURRENT ACCOUNT SURPLUSES:
A THEORY AND EVIDENCE FROM EAST ASIA
Shin-ichi Fukuda
Yoshifumi Kon
Working Paper 13004
http://www.nber.org/papers/w13004
NATIONAL BUREAU OF ECONOMIC RESEARCH
1050 Massachusetts Avenue
Cambridge, MA 02138
April 2007
This is a substantially revised version of the paper "International Currency and the US Current Account
Deficits" that was prepared for the 17th Annual East Asian Seminar on Economics to be held at the
Mauna Lani Bay Hotel, 68-1400 Mauna Lani Drive, Kohala Coast, Hawaii on the June 22-24, 2006.
We would like to thank A. Rose, L. Goldberg, M. Dooley, two anonymous referees, and the other
participants of the conference for their constructive suggestions. The views expressed herein are those
of the author(s) and do not necessarily reflect the views of the National Bureau of Economic Research.
© 2007 by Shin-ichi Fukuda and Yoshifumi Kon. All rights reserved. Short sections of text, not to
exceed two paragraphs, may be quoted without explicit permission provided that full credit, including
© notice, is given to the source.

Page 2

Liquidity Risk Aversion, Debt Maturity, and Current Account Surpluses: A Theory and Evidence
from East Asia
Shin-ichi Fukuda and Yoshifumi Kon
NBER Working Paper No. 13004
April 2007
JEL No. F21,F32,F34
ABSTRACT
The purpose of this paper is to show that macroeconomic impacts might be very different depending
on what strategy developing countries will take. In the first part, we investigate what macroeconomic
impacts an increased aversion to liquidity risk can have in a simple open economy model. When the
government keeps foreign reserves constant, an increased aversion to liquidity risk reduces liquid debt
and increases illiquid debt. However, its macroeconomic impacts are not large, causing only small
current account surpluses. In contrast, when the government responds to the shock, the changed aversion
increases foreign reserves and may lead to a rise of liquidity debt. In particular, under some reasonable
parameter set, it causes large macroeconomic impacts, including significant current account surpluses.
In the second part, we provide several empirical supports to the implications. In particular, we explore
how foreign debt maturity structures changed in East Asia. We find that many East Asian economies
reduced short-term borrowings temporarily after the crisis but increased short-term borrowings in
the early 2000s. We discuss that our results have important implications for the recent deterioration
in the U.S. current account.
Shin-ichi Fukuda
Faculty of Economics
University of Tokyo
7-3-1 Hongo, Bunkyo-ku
Tokyo 113-0033, JAPAN
sfukuda@e.u-tokyo.ac.jp
Yoshifumi Kon
Graduate School of Economics
University of Tokyo
7-3-1 Hongo, Bunkyo-ku
Tokyo 113-0033, JAPAN
ee57014@mail.ecc.u-tokyo.ac.jp

Page 3

2
1. Introduction
In recent literature, it has been widely discussed why the U.S. current account has deteriorated dramatically
during the past decade (see, among others, Obstfeld and Rogoff [2004], Roubini and Setser [2004], Blanchard,
Giavazzi, and Sa [2005]). Although the U.S. current account had been in deficit for most of the periods in the
1980s and the 1990s, its deficits had been almost balanced by Japan’s current account surpluses until the mid
1990s. However, the U.S. current account started to show a dramatic deterioration after 1997 and is now far
from balanced by surpluses of the other industrialized countries (see Figure 1). The first strand of studies
proposes that the recent deterioration in the U.S. current account primarily reflects a decline of the U.S. domestic
saving and an increase in the U.S. demand for foreign goods. The second strand of studies, in contrast, points
out that an increase in the global supply of saving, especially an increase in Asian and Middle Eastern savings,
would help to explain the increase in the U.S. current account deficit. In particular, these studies stress a
remarkable reversal in global capital flows that has transformed emerging-market economies from borrowers to
large net lenders in international capital markets (see, for example, Bernanke [2005], Dooley, Folkerts-Landau,
and Garber [2005], and Caballero, Farhi, and Gourinchas [2006]).
When looking at the recent remarkable reversal in global capital flows, East Asian economies have been one of
the major net lenders after the currency crisis in 1997. Table 1 reports total trade balances of eight East Asian
economies from 1990 to 2004. It also reports their trade balances against the United States and the other trade
partners. It shows that except for Hong Kong and the Philippines, the East Asian economies had trade balance
surpluses in total after the crisis. In particular, except for the Philippines, they have had big trade balance
surpluses against the United States since the crisis and the surpluses have widened in the 2000s. The trade
balance surpluses have been one of the main sources of the U.S. current account deficits since the late 1990s,
especially since the early 2000s.
In this paper, we explore some theoretical and empirical implications of the changed international capital flows
in East Asian economies after the currency crisis. During the crisis, East Asian economies with smaller liquid
foreign assets had hard time in preventing panics in financial markets and sudden reversals in capital flows (see,
for example, Corsetti, Pesenti, and Roubini [1999] and Sachs and Radelet [1998]). Many developing countries
thus came to recognize that increased liquidity is an important self-protection against crises. Among the
strategies for the self-protection, replacing liquid short-term debt by illiquid long-term debt was initially one
popular advice that many economists suggested. However, what most Asian economies have taken more
seriously was raising foreign reserves (see, for example, Aizenman and Lee [2005] and Rodrik [2005]). Foreign
exchange reserves held by developing nations, especially East Asian economies, are now record-breaking, and
stand at levels that are a multiple of those held by advanced countries. The purpose of this paper is to show that
macroeconomic impacts would be very different depending on which strategy developing countries will take for
the self-protection.
In the first part of this paper, we investigate what impacts an increased aversion to liquidity risk can have on
current account and the other macroeconomic variables in a simple open economy model. In the model, each
representative agent maximizes the utility function over time. Since Obstfeld and Rogoff (1997), usefulness of

Page 4

3
utility-based models has been widely recognized. A key feature in our model is that relative size of net foreign
liquid debt to foreign reserve reduces the utility. This is one of the simplest forms that capture costs from holding
liquid foreign debts. At period τ, there is an unanticipated shock that increases aversion to liquidity risk. When
the government keeps the amount of foreign reserves constant, the increased aversion among private individuals
reduces liquid debt and increases illiquid debt. However, because the sum of liquid and illiquid debts does not
change much, its macroeconomic impacts are not large, causing only small current account surpluses. In
contrast, when both private individuals and the government respond to an unanticipated increase of liquidity risk
aversion, the increased aversion increases foreign reserves and may lead to a rise of liquid debt. In particular,
under some reasonable parameter set, it causes large macroeconomic impacts, including current account surpluses
accompanied by depreciation of the real exchange rate.
In the second part of the paper, we provide some empirical evidence in East Asia that supports to the theoretical
implications. In particular, we focus on the changes of foreign debt maturity structures and their implications in
East Asian economies. We find that many East Asian economies temporarily reduced short-term borrowings
soon after the crisis but increased short-term borrowings in the early 2000’s. Since short-term debt is liquid debt,
the former change after the crisis is consistent with the case where only private agents responded to the increased
aversion to liquidity risk. However, the latter change is consistent with the case where the government also
started to respond and accumulated substantial foreign exchange reserves.
Since macroeconomic impacts of the increased liquidity risk aversion depend on which strategy the East Asian
economies take, our results have several important implications. In particular, accumulating foreign exchange
reserves, the U.S. dollar is the dominant reserve currency in the currency compositions. This suggests that
substantial rises in foreign exchange reserves will increase capital inflows into the United States. We point out
that trade account surpluses have been widening against the United States but not against non-US countries in
several Asian economies in the 2000s. Finally, we find that there were substantial depreciations of East Asian
real exchange rates against the U.S. dollar even after the economies recovered from the crisis. We discuss that
the result is also consistent with the model.
There are several previous studies that address determinants of debt maturity structure. For example, Rodrik
and Velasco (1999) argue that international investors with informational disadvantages may choose to lend
short-term to better monitor and discipline borrowers (see also Fukuda [2001] and Jeanne [2004] ). Broner,
Lorenzoni, and Schmukler (2004) argue that emerging economies borrow short term due to the high risk premium
charged by international capital markets on long-term debt (see also Schmukler and Vesperoni [2006]).
However, unlike ours, none of them discussed interactions between debt maturity and foreign reserves that
prevailed in emerging markets in the late 1990s and the early 2000s.
The paper proceeds as follows. Section 2 sets up our small open economy model and section 3 discusses its
implications under constant foreign reserves. Section 4 discusses macroeconomic consequences when the
government chooses foreign reserves so as to minimize its loss function and section 5 presents the simulation
results. Section 6 shows some supporting evidence in East Asia and section 7 considers an implication for the
US current account deficits. Section 8 discusses implications for real exchange rates. Section 9 summarizes

Page 5

4
our main results and refers to their implications.


2. A Small Open Economy Model
The main purpose of our theoretical model is to investigate macroeconomic consequences when the economy
suddenly increased its aversion to liquidity risk. We consider a small open economy that produces two
composite goods, tradables and nontradables. For analytical simplicity, we assume that outputs of tradables and
nontradables, yT and yN, are fixed and constant overtime. Each representative agent in the economy maximizes
the following utility function:

(1) ∑∞
0=
j
j
β
[ U(cT
t+j, cN
t+j) – C (λbA
t+j, Rt+j)], 0 < λ,

where cT
t = consumption of tradable good, cN
illiquid debt, and Rt = foreign reserve. The parameter β is a discount factor such that 0 < β < 1. Subscript t
denotes time period. The utility function U(cT
the disutility function C (λbA
t = consumption of nontradable good, bA
t = net liquid debt, bB
t = net
t+j, cN
t+j) is increasing and strictly concave in cT
t+j and cN
t+j, while
t+j, Rt+j) is strictly increasing and strictly convex in bA
The budget constraint of the representative agent is
t+j.

(2) bA
t+1 + bB
t+1 = (1+rA) bA
t + (1+rB) bB
t - yT - pN
t yN + cT
t + pN
t cN
t + Tt.

where Tt is lump-sum tax, pN
interest rate of illiquid debt. For simplicity, we assume that rA < rB = (1/β) – 1. The assumption that rA < rB
reflects a liquidity premium that makes real interest rate of liquidity debt lower than that of illiquid debt. Since
t is the price of nontradable good, rA is real interest rate of liquid debt, and rB is real
the numeraire is the traded good, the real interest rates and the price of nontradable good are defined in terms of
tradables.
A key feature in equation (1) is that net liquid debt and foreign reserve are in the utility function. In our model,
net supply of domestic debt is always zero, so that bA
t denotes net liquid foreign debt. We assume that relative
size of net liquid foreign debt to foreign reserve reduces the utility. This is one of the simplest forms that capture
potential costs from holding liquid foreign debts. Panics in financial markets and sudden reversals in capital
flows are more likely to happen when the country has higher (net) levels of liquid foreign debts but are less likely
when it has higher levels of foreign reserves. To the extent that ∂C (λbA
Rt+j)/ ∂Rt+j < 0, the function C (λbA
One may interpret the function C (λbA
t+j, Rt+j)/ ∂(λbA
t+j) > 0 and ∂C (λbA
t+j,
t+j, Rt+j) is a reduced form that captures the disutility from such potential costs.
t+j, Rt+j) as a shopping time model where either a decline of bA
Rt saves labor time for reducing liquidity risk. In a closed economy, a fiat money provides such liquidity services
in the money-in-the-utility function model. In a small open economy that has a potential liquidity risk, either a
t or a rise of
decrease of liquid foreign debt or an increase of foreign reserve provides a similar service. In the following