Transcript Some issues for the future 2 Evolution of controls prior to

Capital flows, exchange control
regulations and exchange
rate policy: the South
African experience
``
25 March 2004
1 Introduction
 The
relationship between capital flows,
exchange control regulations and
exchange rate arrangements has long
occupied policymakers
 This paper reviews the SA experience
Figure 1 The South African financial account
1 Introduction
 The
relationship between capital flows,
exchange control regulations and
exchange rate arrangements has long
occupied policymakers
 This paper reviews the SA experience
 Presentation proceeds chronologically:
• Evolution of controls prior to 1983
• The 1985 debt crisis and the reimposition of controls on non-residents
• The liberalisation of controls in the post1994 period
 Some issues for the future
2 Evolution of controls prior to 1983
 The
British influence
 The relationship between controls and
parallel foreign exchange markets
2 Evolution of controls prior to 1983
 We
never thought at the time that we
were instituting a dual exchange rate
system. We thought we were simply
applying exchange control, blocking
funds of non-residents ... there were
some suggestions that we should, in
fact, institute a formal dual exchange
rate system. But we decided against this,
partly because the extended exchange
control was considered to be a
temporary crisis measure (emphasis in
the original). De Kock (1979)
2 Evolution of controls prior to 1983
 The
British influence
 The relationship between controls and
parallel foreign exchange markets
 Evolution of the parallel exchange rate
system:
• the Blocked Rand
• the Securities Rand
• the Financial Rand
The Financial Rand system 1979-83:
 Located
in Johannesburg and London
 Operated via 2 linked channels:
• the ‘cash’ market
• the stock exchanges
 Some implications of the system
• Implications for the BOP
Implications for the BOP:




A non-resident seller of South African assets
had to be matched by a non-resident buyer, with
the exchange rate adjusting to clear the market.
Therefore, non-resident disinvestment had no
impact on the country’s balance of payments.
However, there could be no net investment via
the financial rand either.
To quote Stals (1980) “investments in South
Africa by non-residents with financial rand do
not benefit the balance of payments. The
mechanism only enables non-residents as a
group to shift existing investments in South
Africa from one application to another”
The Financial Rand system 1979-83:
 Located
in Johannesburg and London
 Operated via 2 linked channels:
• the ‘cash’ market
• the stock exchanges
 Some implications of the system
• Implications for the BOP
• Inherent uncertainty regarding duration
 Abolished February 1983
19
79
/0
1/
31
19
80
/0
1/
31
19
81
/0
1/
31
19
82
/0
1/
31
19
83
/0
1/
31
19
84
/0
1/
31
19
85
/0
1/
31
19
86
/0
1/
31
19
87
/0
1/
31
19
88
/0
1/
31
19
89
/0
1/
31
19
90
/0
1/
31
19
91
/0
1/
31
19
92
/0
1/
31
19
93
/0
1/
31
19
94
/0
1/
31
19
95
/0
1/
31
Per cent
Figure 2b The financial rand: 1979-83 and 1985-95
60
50
40
30
20
10
0
Financial rand discount
2 Evolution of controls prior to 1983

“… with the wisdom of hindsight it is clear that
in 1982/3 the monetary authorities were too
optimistic about the financial strength of the
rand and certainly insufficiently sensitive to the
international market perceptions of the basic
weaknesses of the rand, the high liquidity of the
country’s foreign debt, the downside prospects
for the country’s foreign terms of trade, the
doubts concerning its internal monetary
stability as measured by the increase in the
domestic money supply, the low level of real
interest rates, and the low level of fiscal
discipline.” Lombard (1994)
3 The 1985 Debt crisis and the re-imposition
of controls on non-residents
 The
1985 debt crisis: the backdrop
• Between the end of 1980 and the end of
1984, South Africa’s total foreign debt
increased from R12,6 billion to R48,2
billion (US $16,7 billion to $25,5 billion)
• of this, short-term debt increased from
49,1 per cent to 68,0 per cent
• largest share of this was private sector
debt.
 Deteriorating economic fundamentals
and a political catalyst
3 The 1985 Debt crisis and the re-imposition
of controls on non-residents (cont.)
 The
deterioration
of
economic
fundamentals
• Gold price declined from over US$500
per fine ounce in February 1983 to
US$299 per fine ounce in February 1985
• Exchange rate depreciated by 53 per cent
against the USD between September
1983 and January 1985 (44 per cent vs
NEER)
• Weak world commodity markets, 3 years
of drought, excessive money creation in
1983-84
3 The 1985 Debt crisis and the re-imposition
of controls on non-residents (cont.)
 The
•
•
•
•
•
political catalyst
Declaration of State of Emergency 20
July 1985, and subsequent capital
outflows
Rumours banks wouldn’t renew loans
Rubicon speech 15 August 1995
Large-scale outflows until 27 August
when government suspended trading on
the JSE and foreign exchanges
1 September 1985: emergency package
announced including moratorium on debt
repayments and re-imposition of the
financial rand
Was the financial rand system
‘effective’?
 It
depends
 Very little empirical work done
 Can perhaps say something about
insulation provided to interest rates and
exchange rates but this has to be
weighed against the costs of the system
Source: Kahn (2001: 238)
Tests for volatility shifts in nominal
exchange rates:
 Compared
various exchange rates for
first unified (1983-85), financial rand
(1985-95) and re-unified periods (March
1995 – October 1998)
 Conditional
variance
proxies
for
exchange rate volatility in each of the
three time periods obtained from ARCHtype models
 The conditional variance proxies for
exchange rate volatility were then
compared for the various periods
Sample period
st
Sign of
change
Mean
ratio test
Variance
ratio test
1 unified
Finrand
mean
2,287
0,723
-
3,163**
-
standard deviation
3,322
2,805
-
-
1,403**
mean
1,778
0,786
-
2,262**
-
standard deviation
2,757
1,656
-
-
2,772**
mean
1,810
0,602
-
3,007**
-
standard deviation
2,971
1,311
-
-
5,136**
mean
2,453
0,734
-
3,342**
-
standard deviation
3,069
1,342
-
-
5,230**
mean
2,352
0,605
-
3,888**
-
standard deviation
4,923
2,866
-
-
2,951**
Rand/US$
Rand/British ?
Rand/German mark
Rand/Japanese ¥
NEER
Sample period
Finrand 2nd unified
Sign of
change
Mean ratio
test
Variance
ratio test
Rand/US$
mean
0,723
0,531
-
1,362**
-
standard deviation
2,805
1,320
-
-
4,516**
mean
0,786
0,800
+
1,018
-
standard deviation
1,656
1,495
-
-
1,227**
mean
0,602
0,823
+
1,367**
-
standard deviation
1,311
1,103
-
-
1,413**
mean
0,734
1,092
+
1,488**
-
standard deviation
1,342
1,244
-
-
1,164**
mean
0,605
0,696
+
1,150**
-
standard deviation
2,866
1,879
-
-
2,326**
Rand/British ,
Rand/German mark
Rand/Japanese ¥
NEER
Some further evidence


The Engle-Kozicki common feature test
• No evidence was found of a common volatility
(ARCH) process in the dual foreign exchange
rates using the ‘common features’ methodology
Tests for volatility spillovers
• Hamao et al (1990) approach used
• estimated volatility shock for finrand included as
explanatory variable in the conditional variance
equation for comrand, and reverse
• results revealed volatility spillovers from the
commercial rand exchange rate to the financial
rand, but not in reverse direction
4 The post-1994 liberalisation
 The
debate on exchange
liberalisation
 The ‘big bang’ approach
control
• controls provide a disincentive to foreign
investment
• immediate abolition as a signalling device
 The
gradualist approach
• sequencing of financial liberalisation (McKinnon
1973 and 1979, Edwards 1984)
• risks involved in ‘big bang’ approach
 Abolition
of financial rand March 1995
 Further relaxation of controls
4 The post-1994 liberalisation (cont.)
 Resident
institutional investors
• asset swap mechanism introduced July 1995
• involved proposals to swap part of existing
portfolios for the foreign assets of foreign
investors (incorporating measures to “lock-in” the
reciprocal foreign investment)
• some problems, but “… even so they achieved
their basic objective of greater diversification
without capital flight” Vittas (2003)
• 2003: as an interim step towards prudential
regulation, institutional investors allowed to
invest on approval up to the foreign asset limits
 Resident
companies and private
individuals
The foreign exchange and tax amnesty
 Dangers
associated with amnesties
 Motivation: repatriation of capital vs
regularisation of affairs
 Emigrants ‘blocked’ funds to be released
Some issues for the future
 Volatile
capital flows are a fact of life
 Revival of debate on controls with each
crisis
 Softening of attitudes towards controls?
• earlier positions may have overstated
benefits and not emphasised enough the
dangers of liberalisation
• current state of the debate
Some issues for the future
 Implications
•
•
•
•
for SA
Gradual approach seems to be consistent
with the direction debate has taken
obvious that further liberalisation should
seek to maximise benefits and minimise
costs
How is perhaps less straightforward?
Shift away from managing flows directly
and toward limiting the vulnerability of the
economy via prudential regulation is
important, but it would seem there are still
important policy decisions to take
19
79
/0
1/
31
19
80
/0
1/
31
19
81
/0
1/
31
19
82
/0
1/
31
19
83
/0
1/
31
19
84
/0
1/
31
19
85
/0
1/
31
19
86
/0
1/
31
19
87
/0
1/
31
19
88
/0
1/
31
19
89
/0
1/
31
19
90
/0
1/
31
19
91
/0
1/
31
19
92
/0
1/
31
19
93
/0
1/
31
19
94
/0
1/
31
19
95
/0
1/
31
SA cents per US dollar
Figure 2a The financial rand: 1979-83 and 1985-95
600
500
400
300
200
100
0
Financial rand/US$
Commercial rand/US$