Global Liquidity Provision

Olivier Blanchard
Buenos Aires, September 2009
1. Basic facts. Reserves and the crisis
 Global Reserve Accumulation

Emerging Economies: Reserves

(in percent of Global GDP)
14

ROW
12 Emerging Asia
Emerging Europe
Latin America
10 ME/Africa

0
90 92 94 96 98 00 02 04 06 08 Apr. 09
 Currency composition of Global Reserves
(in percent of Global Reserves 1/)

U.S. dollar Japanese yen Euro Others

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
01 02 03 04 05 06 07 08 09Q1
1/ Total allocated reserves.
 Reserves and GDP Declines

Reserves versus 2009Q1 GDP Growth

(Selected emerging economies; in percent)
10
DOM CHN
5 IND
IDN
0
Real GDP Growth 09Q1 v. 08Q3

COL ARG
ZAF PHL PER SER y = -0.0444x - 7.6781
-5 VEN
(two quarter, saar)

CHL R2 = 0.0034

-10 BRA
HUN HRV
THA
-15
MEX MYS
EST
TUR
-20
RUS
-25 LTU

-30
LVA
-35
0 10 20 30 40 50 60
2007 Reserves / GDP
 Reserves against GDP growth, with controls

15

IDN IND CHN

10

ARG
5 ZAF y = -0.0954x + 1.9935
COL
PHL R2 = 0.0264
PER
0
Residuals

CHL
VEN LVA
-5
BRA HUN
EST
THA
-10 TUR
MEX MYS

-15
RUS

-20
0 10 20 30 40 50 60

2007 Reserves / GDP

GDP decliine, controlling for openness, precrisis ca balance, adv manuf share
 Change in Reserves and GDP growth

Changes in Reserves versus GDP Growth

(Selected emerging economies; in percent)
10

DOM CHN
5 IND
y = 0.6321x - 7.5485
IDN
R2 = 0.0727
0
Real GDP Growth 09Q1 v. 08Q3

ARG COL
PER
-5 ZAF PHL
VEN
(two quarter, saar)

SER CHL
HUN
-10 BRA
MYS HRV
-15 MEX
THA
TUR
EST
-20
RUS
-25
LTU
-30
LVA
-35
-15 -10 -5 0 5 10
Reserves (2009Q1-2008Q3) / GDP 2008
 Reserves and Spreads

Reserves and CDS Spreads

(in percent of 2008 GDP)

16 700

14 600
Brazil Reserves

12
500
10
400
8 Mexico Reserves
300
6
Brazil CDS Mexico CDS
200
4 (in bps)
(in bps)

2 100

0 0
Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 Jul-09
 2. Why such high reserve levels?

Byproduct of export led growth? or

Precautionary saving against trade shocks or capital

outflows (private markets can mostly handle the first)?

Depends on the country/region

Twist. What you have is what markets think you must

keep…
 Reserves to Short-Term External Debt
(in percent, average of 2002-07)

600

500

400

300

200

100

0
Baltics Central CIS Latin Emerging China
Europe America Asia (excl.
China)
 3. Are high reserves so bad?

Precautionary saving is highly inefficient against small

probability events (large capital outflows).

Better: Insurance (contingent credit lines) against large

capital outflows.

Why doesn’t it happen? Countries too large; moral

hazard (endogeneity of capital outflows)
 4. Liquidity provision during the crisis.

1. Bilateral credit lines

Fed swap lines. October 2008 (Brazil, Mexico, Korea,

Singapore), 30b each. 6 months, renewed for 9 months

ECB and SNB repos. October, November 2008.

(Hungary, Poland) 5,10b

China. Since September 2008 (Argentina, Hong Kong,

Indonesia, Korea, Malaysia, Belarus)

Chiang Mai. Regional (ASEAN plus Korea, China, Japan)


Bilateral credit lines partly deal with moral hazard,
through choice of country.
Based on economic/political choices.

Effect on reserves? A very partial substitute.

Not a substitute for countries out.

For countries in. Limited duration, explicit and

implicit.
 2. SDRs

Increase in SDR allocation of 250 b (August 2009) (so

not used yet)

Multilateral credit line, available to all members, in

proportion to quotas

Out of 250, (3% of global reserves), 100 to emerging

and developing countries. For example: 3.5 for Korea,
2.5 for Argentina.
 General SDR allocation as percent of June 2009 reserves

0
China India Russia Brazil Mexico Argentina

Available to all members. Unconditional. Does not deal
with moral hazard issue.

Limited size, relative to reserves. Because of moral

hazard, may not want to increase it much more.

Effect on reserves? Close to 1 for 1, but for limited

amounts.
 3. FCL

Contingent credit line. Introduced in March 2009.

Prequalification, no conditionality.

Precautionary. Open for 6 months or 1 year (mid term

review). Commitment fee (25bp for 500-1000% of
quota)

Repayment period 3.5 to 5 years.

Mexico 47b, Poland 21b, Colombia 11b. (not drawn)


Choice of country based on economic criteria. Partly
deals with moral hazard issue.

Potentially large size

Effect on reserves? Because of 0-1 prequalification and

limited qualification period, cannot be sure to have it in
the future: Poor substitute for reserves.
 5. Can we do better? Explorations

FCL looks like a good starting point. Multilateral, partly

deals with moral hazard, potentially large.

Simple idea, building on the FCL. Change 0-1 to a

continuum. How?

Variable premium. Continuous access, at changing

terms.

Premium based on solvency risk, not liquidity risk.

 Are there issues? Yes, many...

Among them:

Effects on liquidity choice. Incentives for countries to

borrow short term. Necessarily bad?

Who assesses solvency risk? Incentives. (EM club?)

Interaction with standard Fund programs?

Two separate windows:
Liquidity window FCL-2.
Program window.
If take program, lower premium.
 Conclusions

Crisis has shown the need for global liquidity provision.

Reserves second best.

Bilateral credit lines second best

FCL provides a useful start. Time to explore how to

improve it.

Connection to regional arrangements. Chiang Mai?

 Reserves and Spreads

Reserves and CDS Spreads

(in percent of 2008 GDP)

16 5000
Argentina CDS
(in bps) 4500
14
Brazil Reserves
4000
12
3500
10
3000

8 Mexico Reserves 2500

2000
6
1500
4 Brazil CDS
Mexico CDS 1000
(in bps) (in bps)
2
500

0 0
Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 Jul-09