Transcript Opportunities and challenges of fluctuating oil

Opportunities and challenges of fluctuating oil
prices for economic development in the context of
the small oil dependent T&T economy
Roger Hosein
1
Rationale
Policy makers in developing economies
are typically very concerned about the
price of oil as wide fluctuations tend to
provide a variety of economic challenges.
For example, volatility in the price of crude
oil can expose an economy to a greater
degree of risk in turn affecting the
investment decisions made by economic
agents.
2
Introduction
This
paper
explores
the
various
opportunities and challenges that arise
from fluctuating oil prices for the Trinidad
and Tobago economy. The discussion is
cast within the framework of the historical
experience of the T&T economy for the
time period, 1966- to May 2009.
3
Statistical Properties of Oil Price Data
For the time period January 2nd 1990
to May 12th 2009, the nominal daily
price of crude oil averaged US$34.25
with a standard deviation of US$23.76.
4
 For this same time period the kurtosis value for
the oil price return series of 23.5 demonstrates
that the data is characterized by a fat tail. The
Jarque Bera test confirms that the returns series
is non-normally distributed. When a data series
has a thick tail it can be modeled using a
conditional normal distribution which implies that
although on any one day the data may be
normally distributed the parameters of normality
change from day to day. The existence of a
small negative skewness statistic suggest the
existence of a left tail.
5
n
04
02
,1
99
0
,1
M
99
ay
1
05
,1
Ju
99
l1
2
3,
Se
19
93
p
19
,1
N
99
ov
4
28
,
Fe
19
95
b
04
,1
Ap
99
r1
7
6,
19
Ju
98
n
25
S e , 19
99
p
06
,2
N
00
ov
0
13
,
20
Ja
01
n
30
A p , 20
03
r1
4,
20
Ju
04
n
24
S e , 20
05
p
06
,2
N
00
ov
6
14
,
20
Ja
07
n
26
,2
00
9
M
ar
Ja
Fig 1: Trends in daily oil price returns and the daily price of oil, January 2nd
1990 to May 12th 2009
160
oil returns
oil price
140
120
100
80
60
40
20
0
6
 When this type of volatility clustering exists it
infers that the variances are autocorrelated so
that shocks in one time period do not die out
quickly but takes time to dissipate.
7
Econometric model
 An EGARCH (1,1) model is preferred to an ARCH/GARCH
specification to capture because:
1) The model does not impose any restrictions on the  ,  and 
2) In contrast to the ARCH/GARCH model, the EGARCH model
allows the coefficient to either have a positive or a negative
value. For | | < 1, the EGARCH (1,1) model is both stationary
and ergodic.
3) The  parameter in the EGARCH model can be used to judge
the impact of shocks. For >0 a positive shock gives rise to more
volatility than a negative shock. The opposite is also true.
4) The  parameter provides an indication of the influence of the
conditional
shock
on
the
conditional
variance.
8
Table 1: Estimates of ARCH, GARCH and EGARCH models
w



LgLikihood
ARCH (1,1)
Normal
0.00046
(0.000)
0.2889
(0.000)
GARCH (1,1)
Normal
0.0000
(0.000)
0.0817
(0.000)
0.9132
(0.000)
GARCH (1,1)
Generalised
error
distribution
0.0000
(0.000)
0.06319
(0.000)
0.9268
(0.000)
11352.8
GARCH (1,1)
t-distribution
0.0000
(0.000)
0.0555
(0.0000)
0.9327
(0.0000)
11388.1
EGARCH (1,
1)
-0.2216
(0.000)
0.1808
(0.000)
0.9906
(0.000)
11217.58
-0.0110
(0.0144)
9

 The coefficient of  is negative and statistically significant and this
infers that for daily oil price returns data for the time period January
2nd 1990 to May12th 2009 as a whole, shocks tended to have
asymmetric effects on oil price returns volatility. Because the sign is

negative, negative shocks reduce volatility
more than positive
shocks.
 Notice that the estimated value of beta is close to unity and is
significant. This infers that when there is a shock to crude oil
volatility, it takes time to die out, i.e. there is some element of
volatility persistence. The failure of shocks to die out provides a
variety of challenges and opportunities for fiscal policy. The rest of
this presentation discusses some of the various opportunities and
challenges that occur with fluctuations in commodity prices for oil
exporting economies.
10
Challenges During oil price Upswings
a) Dutch Disease
Fig 2: The Impact of the Dutch Disease on Employment and Prices
L1NT
L1T
Rm
L0T
Effect
L0NBT
Wages
L0NT
S Effect
W2
W1
W0
0NT
Indirect
Deindustrialization
L1
L0 L2
L0’
L1’
L2’
0T
Labor
11
b) Rent seeking and producer unfriendly institutions
R2
R1
Rents
Profits
B3
B1
B2
P2
P1
0
productive entrepreneurs
rent seekers
0
12
c) Vertical and Horizontal Equity Impact of
GATE
Gate is horizontally efficient but vertically
inefficient (evidence of resource curse behavior).
13
4. Procyclical behavior
Table 2: Correlation between government
expenditure (US$mn) and oil prices, for the
listed time periods.
1966-2008
0.93
1990-2008
0.97
 The values in the Table 2 above show a strong
degree of correlation between government
expenditure and oil prices in the time interval
1966-2008 and for the sub time period 1990 –
2008. This is reflective of a procyclical stance.
14
5. Chasing the goose that lays the golden
egg.
 This procyclical behavioral pattern was compounded by a
type of irrational behaviour on the government behalf which
saw an unfounded increase in the pace at which the speed
of adjustment parameter from an adaptive expectations
perspective.
15
Specifically for the last 3 years the calculated values of the smoothing parameter () are liste
in the Table 3 below:
Table 3: Changes in the adaptive expectations
adjustment parameter, 2006, 2007, 2008.
2006-2007
0.25
2007-2008
0.52
2008-2009
0.90
Source: own computations.
This is a rapid increase in the value of , and has no real foundation. (Note,
the government amidst increased public pressures later modified the budgeted
oil price to US$45).
16
Opportunities during commodity price
upswings
a) Build up of economic resources
 During a commodity price upswing an economy
can develop a tremendous amount of resources
with which to help change its economic
development prospect. The T&T economy has
benefited significantly from the boost in prices of
its comparative advantage products. In
particular, the T&T government was blessed with
an economic windfall between 1999 and 2007 of
US$2,805.6m, with the economic rent for the
period 1999 – 2007 tallying to US$6,955.2m.
17
b) Diversification Strategy
 Resources obtained from the windfall in economic
activity could be used to help enhance the amount of
capital investment made by the government and this in
turn can help to expand both the infrastructural capital
base and also help to diversify the economic base. By
1988 the government of T&T had established four
ammonia plants; it had majority ownership in two of
these and complete ownership in another. It had also
established a steel plant, a granular urea plant and a
methanol plant.
18
C) Social Capital and infrastructural capital
 Significantly the government of T&T spent heavily during the first oil
boom years on building up its human and infrastructural capital. For
example, expenditure on education (an important factor in creating
human capital) increased considerably from US$27.4m to
US$55.7m between 1976 and 1983, an increase of 103%, with
health expenditures (an important factor in maintaining human
capital) increasing from US$0.5m to US$24.4m between 1977 and
1983.
 Concerning infrastructural capital, expenditure on electricity
increased from US$7.8m in 1976 to US$17.3m in 1980 with housing
expenditure increasing substantially from US$7.8m to US$191.8m
during the interval of time 1976-82. Other important infrastructural
capital areas such as roads, national transportation, water and
telecommunications also benefited substantially, as indicated in
Table 4.
19
Table 4: Funds Allocated to Long-term development projects in T&T, ($USm), 1976 – 1983.
Education
Health
Electricity
Housing
Roads
National Transport
Water
Telecommunications
7.80
0.47
6.11
9.11
7.24
1.93
1976
27.39
1977
19.62
0.50
9.11
6.67
19.36
9.28
26.46
13.70
1978
22.20
1.22
8.07
37.09
22.88
13.74
64.58
18.66
1979
27.58
6.82
9.27
49.45
55.85
9.12
78.19
25.76
1980
17.23
4.89
15.92
86.05
98.91
22.23
44.76
29.17
1981
27.37
10.07
17.34
111.18
16.09
13.22
45.24
29.57
1982
47.39
13.13
3.61
195.10
189.48
8.64
34.98
41.81
1983
55.70
24.35
191.80
87.67
Source: Accounting for the Petrodollar (1983).
20
d) Other interventions
 Apart from these direct long-term development
investment outlays listed above, a plethora of
subsidies were granted. For example, between
1973 and 1983, the basic food subsidy
increased from US$1.6m to US$92.5m with
subsidies to the utilities increasing by 3,351.1%
during the same interval of time from US$10.7m
to US$369.3m. No doubt, these trends in
government expenditure and government
subsidies helped to alter the factor endowment
bundle of the T&T economy.
21
e) FDI
 The type of foreign investment a small open developing
economy like T&T attracts is cost based. During a
commodity price upswing the profitability of crude oil
production expands and if the increase in the price level
is expected to span the medium run in length then this
can enhance capital formation in the energy sector.
During the period 1995 to 2008 the correlation between
FDI inflows and the amount of meters drilled for crude oil
was 68%. FDI inflows provides opportunities along the
lines proposed by Sir Arthur Lewis for the benefiting host
economy.
22
Table 5: Trends in some aspects of the Factor Endowment Portfolio of T&T, 1966 - 2007.
Year
Physical capital
Human
capital
Physica
l labour
Labour
force
(000)
US$ Gross
Capital
Formation
(m)
US$
FDI m
Primary
school
graduate
s
1966
166.9
24.85
216063
351
1970
212.5
83.20
227254
364
1975
594.6
180.56
204004
1980
1908.5
143.42
1985
1384
1990
Infrastructural Capital
Water: annual
production in
millions of
gallons
Electricit
y mn
kilowatts
hours
crude oil
production
(m barrels)
19385.8
701
55603
21422.3
907.3
51048
391
23331.1
1124.2
166763
431
32269.1
49.67
168790
474
1181.1
699.9
109.40
189752
467.6
1995
847.1
295.7
191641
2000
1657..3
679.5
2001
1838.9
2002
2001.8
2003
2987.7
2004
Entrepreneurs
Natural gas
production
(m of cubic
metres)
Graduates with
first degrees from
the U.W.I. St
Augustine Campus
3367.6
100
25692
3428
133
78621
63525
3580.2
288
1892.9
77618
60510
5601
388
50795.5
2903.1
64259
49823
7550
433
164.9
56166.4
3466.1
55039
41672
6645.5
547
521
205.4
55051.1
4228.6
47576
36746
7757.4
839
168734
564
290.6
65759
6307.7
43680.5
34710
15473.6
1661
834.9
163206
576.5
306.9
64044
5688.2
41521.3
32151
16485.8
1633
790.7
147328
586.2
318.2
73288
6036.1
47706.6
38792
18861.6
1708
583.1
141401
596.6
319.8
72591
6423.9
48981.1
39992
26794
1817
2317.7
972.7
136,327
613.5
321.2
76135
6709.6
44984.7
35898
30255
1952
2005
2458.0
598.7
134298
623.7
327.6
77562
6803.2
52739.6
43879
33250
2006
2592.5
512.2
133562
625.2
331.2
79231.1
7102.4
52104.8
2007
2944.3
830
622.4
Telephone
lines in
service 000
connections
Land Resources
35.1
43807
marine
production
000 barrels
40057.8
41906.78
23
2008
Opportunities during a downswing in
commodity prices
a) Improvement in fiscal accounting process
During an upswing there is an element of
crowding out of the amount of non oil
revenues collected as a proportion of total
government revenues. Although ideally
the government should have made these
changes during a commodity upswing,
during a commodity downswing they are
forced to set an appropriate target for the
non oil fiscal account balance.
24
b) Worker Ethic
 During a boom, worker ethic and wages may have become
unrealistic. A good example here is the Community and
Environmental Protection and Enhancement Program (CEPEP)
which was introduced by the T&T government in 2002. Basically the
program provides employment for unskilled and semi-skilled
persons. The challenge with CEPEP, however, is that it needs to do
more to equip workers with a greater stock of skills to enhance their
long run employment prospects. As it stands the work is light and
productivity is very low. During a downswing in commodity prices,
there is the opportunity to target the restoration of worker ethic and
any misalignment between worker productivity and wages may be
corrected. The economy may also have realized a greater
dependence on government handouts and transfers and subsidies
and this may now be corrected. A decline in commodity prices gives
government an adequate block of reasons to reduce its hand out
mentality, and to promote a better work ethic amongst its population.
25
Challenges during a commodity price
downswing
a) Public Perception
 Meeting public expectations in terms of social
handouts and cutting public expenditures in line
with expected revenues is one of the main
challenges that a commodity exporting economy
may face in periods of declining commodity
prices. This type of action may sprout social
tensions. For the T&T economy, transfers and
subsidies as a percentage of total GDP had a
88% degree of comovement with the trend in oil
prices. Specifically whilst in 1994, transfers and
subsidies as a percentage of GDP was 7.4%,
this increased to 12.4% by 2008.
26
b) Inflation
One would expect that during a recession
the rate of inflation would decrease as the
level of aggregate demand moves into
deficit. However, during the first recession
that engulfed the T&T economy in the time
period 1983-1993, the average rate of
inflation was 10%.
27
c) Economic Growth
Economic growth can potentially
be compromised. This was the
case during the 1980s.
28
d) Unemployment
For the T&T economy, there is an inverse
relationship between the current account
balance and the unemployment rate.
29
e) CAB and M/L Condition
 One major challenge in oil abundant economies is there
may arise the need to devalue the currency and this is
sometimes done without a proper understanding of the
Marshall - Lerner condition and whether that condition
holds. Thus the T&T economy, in the midst of declining
current and fiscal account deficits during the 1980s
engaged a 33% devaluation of the T&T dollar in
December 1985 and a further 15% devaluation in August
1988. In 1993 the T&T government floated its dollar and
associated with this float was a 25% depreciation of the
T&T dollar. Between 1985 and 1993 the T&T dollar
therefore lost 54% of its value to the US$ and the current
account balance moved from US$-105.3 to US$-107.8m.
30
Minimizing oil price volatility effects
on the macroeconomy
a) SF and the Long term oil Price
 When oil revenues are siphoned off into a SF this limits
the extent of the volatility that is transmitted. Whilst a
stabilization fund does exist in T&T it holds a mere
US$3bn. Clearly then a critical policy adjustment is that
the SF has to be complemented by a long run price of
crude oil. Windfall prices should be defined as prices
above this long term level, and relevant surpluses should
be creamed off to a stabilization fund. To calculate a
long term price of crude oil, one could possibly use a
Hodrick Prescott filter and extract the long term trend
component from the real oil price series and then
smoothen it using some technique.
31
Table 6: Estimated windfall per barrel of crude oil using a
long term price of US$35
Actual Price
of oil
APBudgeted
Price
AP-Long
term price
2004
41.5
16.5
6.5
2005
56.5
31.5
21.5
2006
66
33.2
31
2007
72.3
22.3
37.3
2008
100
30
65
Source: computed
32
b) Non Oil Balance
 Another policy requirement for the T&T economy to consider is to enhance
the resilience of its economy to oil price shocks. This model requires a
strengthening of the economic base away from a narrow dependence on
hydrocarbon based resources. This strategy can involve a clear non oil
revenue based policy stance. The diversification of the asset base of the
economy would require prudent management as the BT sector expands
over time. In this regard the government planned industrial investments are
welcome but these has to be founded on comparative advantage theory.
The government has to be wary that its attempts at picking winners does
not simply result in rent seeking and inefficiency.
c) Governance Process
 There is a need to strengthen the general governance process. The extent
of corruption in the T&T economy has started to become too entrenched
and there is need for a greater degree of public accountability. Recent
empirical literature such as Collier (2007) have demonstrated that countries
with strong institutions tend to be less affected by the resource curse than
others. In 2001 T&T’s corruption perception index was 5.3, but by 2007 this
had declined to 3.4, indicating a greater degree of corruption. A strong
thrust must be made to review this situation.
33
Thank you
34