Events European Course

1
Country Risk and Capital Flows The case of South
Africa by group no. 2
2
Background

• South Africa has undergone a great deal of
  political change
• The implication were among a number of thinks a
  significant reduction in the country risk
  associated with investments in SA.
• Implementation in the GTAP model
• standard model and GE closure
• no need for swaps

3
Implementation

• Shock
• Estimate based on real observations
• Measure based on Deutchmark bond issues fell by
  13.5 per cent
• We shock the cgdslack by this amount (-13.5
  per cent)

4
Model equations
Expected rate of return rore(r) rorg
cgdsslack Allocation of investments
rore(r) rore(r) - RORFLEX(ke(r)-kb(r))
5
Results

6
Results - demand for intermediates and factors

• Derived direct demand from cap. good sector
• Techmnfc
• Svces
• Factor market effects
• factor intensities in expanding sectors
• factor prices ()
• Supply prices (zero profits)
• Trade effects (effects on other regions)

7
Risk Premium decline FTA among SA and Rest of
Southern Africa

• Pre-shock data FTA and Non-FTA
• Post-shock data FTA and Non-FTA
• Welfare effects on Rest of S. Africa

8
Pre-shock trade flows RESTSA - SA (US m)
9
Pre-shock inputs to capital goods industry, SA
10
Post-shock Welfare effects
11
Post-shock Increased Allocative Efficiency, Rest
of Southern Africa
12
Small Group II

• Extension
• Effects of Changing RORFLEX parameter

13
Introduction The RORFLEX and The Rate of
Return ApproachThe sensitivity of the RORFLEX
Backward shock
?
??
???
14
Investment depends on expected rate of return
in next period
This rate declines as capital stock rises..
The rate at witch this decline is expected is a
function of the flexibility parameter RORFLEX(r )
gt 0 So
RORE( r )RORC( r)KE( r ) / KB( r ) -RORFLEX(
r )
15
More about RORE( r)...
ROR Flexibility(parameter)
Expected rate of return
End of period Capital stock
RORE( r )RORC( r)KE( r ) / KB( r ) -RORFLEX(
r )
Beginning capital stock
Current period rate of return
Thanks, to Soren and Rob
16
Modeling risk

• The global bank equalizes risk-adjusted rates of
  return, so that risk-adjusted rates for all
  regions are egal to some global average
• with
• RORE( r) non risk-adjusted expected rate of
  return
• RISK( r) ratio of equilibruim returns in region
  r to the global average rate of return

17
A risk premium
We have RORE(r )/RISK( r)RORG gt RORE( r)
RISK( r) RORG by total differentiation rore(r
) rorg risk( R) We have also rore(e) rorg
cgdslack(s ) equation 11 , Hertel and
Tsigas so cgdslack(r )risk(r ) gt
cgdslack is equivalent to the percentage change
in the variable RISK We can refer to this,
rather than risk
18
RORFLEX and regional investment changes

• We assume that investors behave that changes in
  regional rates of return are equalized across
  regions
• the global rate of return changes by the same
  percent

?

• A small RORFLEX needs large changes in end of
  capital stock, KE(r ), to induce small changes in
  RORE(r )
• THEN
• Low values of RORFLEX(r ) lead to big changes in
  regional investment
• High value lead to small changes In this case
  the supply of new capital goods is not very
  sensitive to changes in the expected return

19
THE SENSITIVITY OF THE RORFLEX
ROREf RISK(r)
RORFLEX
-13
Cgdslack(r )
5
10
20
(Supply Capital)
dK
20
Impacts of RORFLEX

• Impacts on net capital inflow and welfare
  effects.
• Simulations change value of RORFLEX
• Shock is always the same (cgdslack -13.5)
• RORFLEX 10 (base run)
• RORFLEX 5
• RORFLEX 20
• RORFLEX is changed for all five regions.

21
Net Capital Inflow
22
Welfare Effects
23
Backward shock
Net Capital Inflow in SA
15676
5726
Time
shock -13.5
24
Why an Adjustment?
Net Capital Inflow in SA
5726
5027
-414
-1797
Time
25
Possibilities of Adjustment

• Problem Change of Investment is too big compared
  with the historical Values
• Possibilities
• Shock Assumption is still the same (Shock
  cgdslack13.5)
• RORFLEX Increase of RORFLEX reduces Investment

26
Calculated Value for RORFLEX
5441
13.25
27
EXTENSIONAL
SIMULATION

• GROUP 2
• SINICHI YAMAGUCHI
• MANABU SHIMASAWA

28
How to shock ?
SHOCK BASE SHOCK capital
inflow into South Africa negative shock to
cgdslack cgdslack("safrica")
-13.5 ADDITIONAL SHOCK
technical progress at TECHMNFC SVCES positive
shock to ao
ao("TECHMNFC","SAFRICA") 1.0
ao("SVCES","SAFRICA") 1.0
29
Initial effects induced by Tech change
Capital inflow Direct Investment
Capital Stock increase
Management resource ( know-how , hi-tech etc)
increase
Technical Progress
Price fall (equation (6))
Output increase (equation(35)(36))
30
Welfare Decomposition (1)
A summarized welfare report
31
Welfare Decomposition (2)
C1 TECH decomp. Tech change
C11 AO cont. output augm. tech change
32
Market Price (pm) change
change
33
Volume change in Endowments
BASE
EXT
34
Export quantities change
change
35
SOME CONCLUSIONS
Additional impacts on ao( , ) cause
SVCES sector growth endowments
concentration to SVCES Other sectors
scale decrease Total export decrease
(more capital inflow)
36
Endogenous capital stockGregg Watts andA.
Salazar Brandao
37
K long run capital stock
38
Variable Original Endogenous Capital Stock Gross
investment 36 US b 30 US b Depreciation 19
US b 22 US b Net Investment 17 US b 8 US
b Capital Stock 0 18 Rental
price 7.6 -8.5 qgdp 0.6 6.1
39
Original Endogenous Capital Stock Bal. of
Trade -10 US b -1 US b Welfare 3 US b 8
US b cnt_tot 1.9 US b -0.3 US b pexport
(share) 99 100 ------------------------------
--------------------------------------------------
Perc. changes Exports Price Qty. Price Q
ty. Extraction 3.0 -13 -0.2 0.7 Hvymnfc 6.3 -
21 -1.6 6.4 Svces 7.0 -18 -0.8 2.2
40
Perc. changes Original Endogenous Capital
Stock Price Qty. Price Qty. Capital
8 0 -9 18 Land -12 0 15 0 Nat. Res, -13 0
6 0 Unsklab 8 0 4 0 Sklab 10 0
4 0 ----------------------------------------------
---------------------------------- Agriculture 4.
9 -3.9 -0.4 3.3 Hvymnfc 6.3 -8.4 -1.6 7.2 Sv
ces 7.0 2.7 -0.8 6.0