Sylvia Piterman
Director, Foreign Currency Department
27 April, 2000
Liquidity as a Safety Net1
- A few comments on the concept of liquidity risk -
Introduction
The purpose of this panel is, of course, to discuss the management of liquidity risk. But before getting
into the intricacies of the subject allow me to emphasize that the multitude of vulnerabilities common
to transition economies necessitates maintaining several lines of defense. This conference
concentrates on one of them - the third. The first line of defense would include all the elements of
prudent macro-policies: a proper fiscal policy and an appropriate monetary policy, combined with an
exchange rate regime as flexible as possible. The second line of defense consists of a sound financial
system, and especially a strong banking system.
The third line of defense is the proper management of the country's international debt and liquidity.
Since the first signs of weakness of a national economy always emerge in the foreign exchange
market, maintaining a strong foreign liquidity position is important in assuring the economy's good
health. This line of defense, to which increasing attention has been paid in the aftermath of the recent
financial crises in emerging markets, will be the subject of the rest of my comments.
Alan Greenspan introduced the concept of “liquidity at risk” in the World Bank Conference on
Reserves Management a year ago. According to his definition, liquidity at risk can be calculated by
considering the “country’s liquidity position under a range of possible outcomes for relevant financial
variables (exchange rates, commodity prices, credit spreads, etc.)” and the probabilities of these
outcomes. A country can manage its liquidity at risk position by different combinations of reserves,
debt and their maturity. “For example, an acceptable debt structure can have an average maturity in
excess of a certain limit. In addition, countries could be expected to hold sufficient liquid reserves to
ensure that they could avoid new borrowing for one year with a certain ex-ante probability, such as 95
percent of the time.”
Mr. Greenspan mentioned setting international standards for the level of liquidity at risk, which can be
handled in different ways. One possibility is to increase the level of reserves through mobilizing
long-term debt. Another option would be “a wide range of innovative financial instruments –
contingent credit lines with collateral such as the one maintained by Argentina, options on commodity
prices, opposite options on bonds, etc.”.
The concept of liquidity at risk deserves careful consideration, and I would like to address five issues:
Is it possible to calculate liquidity at risk? How can the desired level of liquidity at risk be attained?
What can be achieved by maintaining the desired level of liquidity, and what cannot? Is this concept
at all applicable to countries with flexible exchange rate regimes? What can be said about the Israeli
experience?
Calculating liquidity at risk
Liquidity at risk can be calculated by estimating the country’s liquidity position for a range of possible
values of relevant financial variables and the probabilities of these scenarios. Doing this is not a
simple exercise and the necessary database is not always available. But keeping in mind that the
concept of “Value at Risk” is already quite familiar, it might be possible to develop and to implement
the approach of “liquidity at risk”. Moreover, in recent years we have seen a substantial development
of early warning models. The database needed to calculate liquidity at risk includes the same variables
as those used in these models. Perhaps the IMF would find it beneficial to do the needed calculations
and provide them to member countries (one of the early warning models was developed in the IMF
Research Department). An obvious advantage of the IMF calculations is that the same methodology
would be used for all countries. Such results could be helpful for the IMF as well in analyzing the
liquidity position of different countries.
Adjusting to the desired level of liquidity at risk
Let us assume that the authorities wish to improve their liquidity position. How can this be achieved?
In the long run, policy will have to be aimed at improving the underlying performance of the
economy, by reducing the current account deficit, encouraging balanced growth, especially in exports,
promoting adequate domestic savings, and so on. In the medium term, it might be feasible to improve
liquidity by long-term government borrowing aimed at building up foreign reserves, by changing the
structure of the government external debt - increasing the weight of long-term debt – or a combination
of both. But changing the maturity structure of external debt takes time and entails costs, especially if
a country finds it difficult to raise long-term debt. Therefore, any application of such a rule has to be
gradual. It is not advisable to improve the liquidity position of the economy by bolstering reserves
through market intervention, since this would necessitate overly restrictive fiscal and monetary
policies, aimed at attracting capital inflows and creating a surplus in the current account.
Establishing contingent credit lines can also help achieve the liquidity at risk target. At first sight it
looks like a perfect solution for countries that need to strengthen their foreign position. However, on
further thought, lines of credit look just a little too perfect: can such lines be big enough to be of
consequence when a real crisis hits? Will the international banking system refrain from defending
itself through cutting other credit lines to a country (to the private sector for example), when the
contingent credit lines are used? Will such contingent arrangements not be too big of a temptation for
governments not to pose the familiar moral hazard problem? Moreover, it is not clear that the cost of
maintaining contingent credit is lower than borrowing long-term and maintaining short-term reserves.
Such reservations are by no means an indication that this idea is a bad one. It is, after all, an insurance
arrangement, and it is believed that in general insurance does work. But only time and further
experience will tell how effective and efficient contingent credit lines really are.
What can be achieved by maintaining a liquidity at risk target and what cannot?
Adopting a target for liquidity at risk can provide a cushion through increasing the financial stability
of the economy. A stronger financial structure of assets and liabilities increases the credibility of the
economy, enhances its access to international capital markets and can substantially reduce the odds of
a crisis taking place.
Nevertheless, achieving some target level of liquidity at risk cannot always provide sufficient
protection against shocks. In the case of a breakdown of confidence in the government’s policy,
demand for foreign exchange might be much larger than the amount of reserves and credit lines,
which were determined according to the above mentioned criteria. Since globalization is increasing, a
strong financial situation according to historical standards may not be adequate for present flows.
Demand for foreign currency might come from different directions and in amounts much larger than
in the past.
Is the concept of liquidity at risk relevant for countries with flexible exchange rate regimes?
The concept of liquidity at risk was developed against the background of a managed exchange rate
regime of one type or another, since such regimes make liquidity essential. However, an economy
with a floating exchange regime, which has not yet achieved proven long-term stability, will also need
huge liquidity. Otherwise, a floating exchange rate regime can bring about extreme nominal
instability.
Therefore, when the exchange rate is floating, the level of reserves should be set at a level that will
prevent extreme fluctuations in the exchange rate – and in the price level - anot according to the
liquidity needed for intervention in different scenarios. This can be achieved by estimating the
minimum level of reserves that ensures access to international capital markets for the economy and
thereby contributes to achieving some limit for “value at risk”. “Value at risk”, in this case, refers to
the exchange rate and the price level fluctuations.
It should be noted that adopting a rule for the optimal level of reserves - when the Central Bank does
not intend to use them - might appear to be a very costly strategy. It is clear that borrowing will
generally be more expensive than investing, all the more so as the maturity of the debt is significantly
longer than the maturity of the assets. Moreover, does holding liquid assets make sense at all, while it
is possible to borrow in the deep international financial markets, should the need arise? I think that the
exchange rate regime per-se is not of crucial importance for setting the optimal level of reserves. It
seems that the main factor is the extent of proven long-term stability. Access to the capital markets
cannot be taken for granted, and capital might not be available precisely when it is most urgently
needed. Moreover, holding high levels of reserves can lower the cost of borrowing for the economy as
a whole, making the strategy optimal for the country.
The Israeli experience
The concept of the optimal level of reserves in general, and in particular for Israel, has evolved over
time. Fifteen years ago “three months of imports” was considered the optimal level of reserves for
Israel. Research conducted in the BoI at the end of the eighties introduced some sophistication. It took
into account the volatility of capital flows, “window dressing” effects and other factors 2, to find the
optimal level of reserves.
Currently the exchange rate is floating within a very wide band. Nevertheless, it is much more
difficult nowadays to predict the volatility of capital flows. Moreover, given the financial turmoil of
recent years, it seems that the optimal level of reserves today is much higher than it was at the
beginning of the nineties.
The current level of reserves is close to 21b$. As can be seen in Table 1, the level of foreign reserves
has grown in comparison to all major financial variables. For example, their average level in 1999 was
larger than the level of short-term liabilities of the country and exceeded the level of BoI domestic
liabilities to the private sector.
How did the improvement in the liquidity position take place and what were the forces behind it? In
the case of Israel, debt had already been of a long maturity for many years; but reserves increased
dramatically during 1995 to 1997 as a result of heavy intervention, aimed at avoiding the exchange
rate impact of huge private sector capital inflows. Therefore, this improvement of the liquidity
position was a by-product of a tough monetary policy coupled with the constraints of the prevailing
exchange rate regime.
As mentioned before, the exchange rate is currently floating within a very wide band (its width is 37%
- relative to the average of its boundaries - and it is increasing by 4% a year). The Bank of Israel has
committed itself not to intervene in the foreign exchange market as long as the exchange rate remains
within the boundaries set by the band. Some commentators fault what they perceive to be the central
bank's failure to use the stock of foreign reserves in order to stabilize the exchange rate. But more to
the point, questions are raised concerning the need for such a high level of reserves if they are not
'used' by the central bank.
Table 1
Israel’s Foreign Currency Reserves relative to various aggregates (%)
(Averages)
|
Months
of
Imports
|
Gross
Ext.
Debt
|
Short
Term
Ext.
Debt
|
Foreign
Currency
Credits
|
M1
|
Unlinked
Shekel
Assets
|
The
Monetary
Base
|
The
BOI’s
Shekel*
Liabilities
|
1991
|
3.2
|
21
|
59
|
76
|
223
|
64
|
328
|
--
|
1992
|
2.7
|
18
|
55
|
67
|
177
|
50
|
283
|
--
|
1993
|
2.1
|
15
|
44
|
62
|
145
|
37
|
219
|
--
|
1994
|
2.1
|
15
|
43
|
66
|
142
|
33
|
208
|
--
|
1995
|
2.5
|
19
|
50
|
62
|
183
|
34
|
295
|
139
|
1996
|
2.6
|
21
|
49
|
55
|
190
|
32
|
288
|
107
|
1997
|
4.6
|
33
|
84
|
75
|
316
|
48
|
466
|
110
|
1998
|
5.9
|
39
|
108
|
82
|
392
|
55
|
520
|
111
|
1999
|
5.8
|
39
|
103
|
77
|
392
|
50
|
506
|
105
|
* Includes the monetary base, time deposits of commercial banks and government deposits connected
to outstanding balances of short term loans (Makam).
The answer to this legitimate concern is that high reserves are doing their job by just being there.
Precisely this very level of liquidity was one of the factors that protected Israel from the financial
turmoil in 1997 and 1998. It helped to maintain a relatively low risk premium on its foreign debt and
to maintain relative exchange rate stability without any intervention by the Bank of Israel.
A few concluding remarks
Maintaining an appropriate level of liquidity is an important line of defense for countries in transition.
It cannot completely prevent currency crises but it can decrease the probability of its occurrence by
enhancing the economy’s access to the international financial markets.
The concept of liquidity at risk can help establish the appropriate level of liquidity. However, the right
level is not easy to calculate and some work by the IMF, which collects all the relevant data, can be of
great help. Converging to the desired level of liquidity at risk has to be done gradually. Contingent
credit lines might be of value but more experience is needed to assess their robustness.
The concept of liquidity risk was developed for countries with managed exchange rate regimes.
However, countries in transition maintaining floating exchange rate regimes also need substantial
liquidity, if they aspire to avoid extreme nominal instability. One way to establish the optimal level of
reserves in this case can be through setting some target for “value at risk” and attaining the liquidity
position that is necessary to achieve this target.
Israel is an example of a country with an almost totally floating exchange rate regime that maintains
large reserves. My impression is that the strong liquidity position of the country has helped to protect
Israel from the financial turmoil of 1997 and 1998.
-------------------------------------------
1Prepared for the Conference of the World Bank Treasury on “Current Issues in Reserve Management for Central
Banks in Europe, Middle East and North Africa”, Istanbul, Turkey, May 4-5, 2000.
2See Ben-Bassat, A. and Gottlieb, D., “Optimal International Reserves and Sovereign Risk”, Journal of
International Economics 33, 345-362, 1992 and Ben-Bassat. and Gottlieb, D., “The Effect of Opportunity Cost
on International Reserve Holdings”,Review of Economics and Statistics 74, 329-332, 1992
|