72 R.W. Faff, T.J. Brailsford Journal of Energy Finance and Development 4 1999 69–87
industry equity returns to an oil price factor. The paper is organized as follows. Section 2 discusses the importance of oil across the different Australian economic sectors. Section 3
outlines the data and research design, and the results are presented in section 4. Section 5 briefly discusses results from some sensitivity and further analysis. The final
section presents the summary and concluding remarks.
2. Oil and the Australian economy
Australia has virtually been self-sufficient in petroleum products over the last 20 or more years. Despite this fact, Australia has engaged in import and export activity
of petroleum products. Generally, during the 1980s Australia was a net importer, whereas during the 1990s, Australia has been a net exporter.
1
However, the net external trade has typically been less than 2 of domestic consumption.
Given Australia’s vast geographic size, relatively small and diverse population, and its remoteness, the costs of transportation and freight potentially constitute a major
component of the costs of many Australian companies, and the price of oil is likely to have an impact on these costs. Conversely, some Australian industries derive
considerable revenue from oil and oil-related products and hence changes in oil prices will affect the profitability of these industries.
It is difficult to formulate predicted signs and relative magnitudes across specific industries of oil factor sensitivities. Industries with a relatively high proportion of
their costs devoted to oil-based inputs, such as Transport, are expected to have a negative sensitivity. Conversely, in the absence of offsetting effects, we would expect
a positive oil return sensitivity in oil and oil-related industries, in which oil directly impacts the revenue side of the income statement. However, in general, the impact
of oil price changes on equity prices will depend on the ability of firms to pass on the effect to customers through changing goods prices.
Moreover, firms can protect themselves against adverse movements in oil pricing through hedging using derivative instruments. For instance, airline companies are
likely to enter into energy futures or longer-term oil delivery commitments. However, the extent to which hedging occurs will make the sensitivity to oil price changes
harder to detect and bias against significant findings. Furthermore, given the growth in derivative products and the improved understanding of risk management, we expect
hedging practices to have become more common in recent years and hence the sensitiv- ity to oil price changes, if any, will have weakened over time. But ultimately, the
extent of an “extra-market” sensitivity to oil price returns is an empirical matter.
In an attempt to form predictions about the sign of the oil factor sensitivity, we first group firms into industries following the Australian Stock Exchange ASX
Industry Classification Report 1997. We initially hypothesize that there are four industries in which oil price changes are expected to have a net impact on revenue
side.
2
These industries are 1. Gold sub-industry code 14: Gold, Oil.
2. Solid Fuels sub-industry code 36: Coal, Oil.
R.W. Faff, T.J. Brailsford Journal of Energy Finance and Development 4 1999 69–87 73
3. Oil and Gas sub-industry code 41: producers, sub-industry code 42: explorers, sub-industry code 43: investors and sub-industry code 44: distribution.
4. Diversified Resources sub-industry code 52: Oil, Steel, Mining, sub-industry code 55: Coal, Gold and Oil and sub-industry code 56: Oil, Gold, Investment.
Conversely, in the absence of offsetting effects, we expect a negative oil return sensitivity in the non–oil-related industries, wherein oil price changes directly impact
on costs. Specifically, we expect the potential for a negative oil-price sensitivity is greatest in industries with a relatively high proportion of their costs devoted to oil-
based inputs, such as Transport.
3
To assess these effects, we have consulted the Australian Input-Output Tables for 1993–1994, and the relevant details are reported
in Table 1. Table 1 reveals that the Australian Bureau of Statistics broadly classifies Australian
industries into 35 different sectors column 2. Unfortunately, this scheme does not readily translate to the industry classification system employed by the Australian
Stock Exchange ASX, which uses 24 categories, for which we have available equity price data.
4
The final column of Table 1 is an attempt to link the ASX industries to their ABS industry counterparts. In some cases, the match is quite tight; for example,
ABS industry 12 and ASX industry 10 are both labeled “Chemicals.” In other cases, the relationship seems less than ideal, or at least affected by the mixing of other areas
of activity. For example, ABS 7 Textiles is matched with ASX 22 Miscellaneous Industrials on the basis that ASX sub-industry 223 is labeled “Textiles” ASX, 1997,
p. A2. However, a problem arises with this match because ASX industry 22 is a “Miscellaneous” industry as it also contains Automotive 222, Household Durables
224, and other sub-industries. Consequently, the matching of ABS and ASX indus- tries provided in Table 1 needs to be treated with due caution.
The major aim of Table 1 is, using the extracted Input-Output Table data, to provide an indication of the differential importance of oil prices on the costs of Australian
industries. These data are used to calculate Direct Requirement Coefficients DRC for the ABS industries. According to the ABS Australian National Accounts Input-
Output Tables 1993–1994 p. 8, the DRC are obtained by calculating inputs as a percentage of the output of an industry and can be used for estimating the input
requirements for any given output of that industry. To make the DRC information more easily interpretable across industries, we report in the second-to-last column of
Table 1 the relative DRC RDRC, which is calculated as the ratio of a particular industry’s DRC to the average DRC across all industries. For example, the Transport
Industry ABS 26 and ASX 14 has a high RDRC, at almost five times the economy- wide average. Perhaps surprisingly, the maximum value of the RDRC is in the ABS 2
Forestry and Fishing industry ASX 12 Paper and Packaging. There are five other ABS industry classifications for which the RDRC value exceeds unity. These are ABS 1
Agriculture and Hunting ASX 22 Miscellaneous Industrials, ABS 3 Mining ASX 2 Other Metals, ABS 12 Chemicals ASX 10 Chemicals, ABS 15 Basic Metals and
Products ASX 2 Other Metals, and ABS 35 Personal and Other Services ASX 21 Miscellaneous Services. For these industries, other things being equal, we may predict
a negative sensitivity to the oil price factor. However, the Miscellaneous Services
→
R.W. Faff,
T.J. Brailsford
Journal of
Energy
Finance and
Development
4 1999
69–87 Table 1
Direct Requirement Coefficients across Australian Industry Classifications Petroleum
Direct Coal
Australian Requirement
Products Production
Coefficient Relative
ABS Industry Classification
a
AUDm AUDm
DRC
b
DRC
c
ASX Industries ASX Industry number 1 Agriculture; Hunting
632.4 22,151.6
0.02855 2.90609
Miscellaneous Industrials 22 2 Forestry Fishing
171.5 2,725.7
0.06292 6.40484
Paper Packaging 12 3 Mining
552.6 31,047.2
0.01780 1.81180
Other Metals 2 4 Meat Dairy Products
62.3 15,386.7
0.00405 0.41216
Miscellaneous Industrials 22 5 Other Food Products
131.0 15,645.4
0.00837 0.85233
Food Household 9 6 Beverages Tobacco Products
34.9 6,222.3
0.00561 0.57095
Alcohol Tobacco 8; Entrepreneurial Investors 18
7 Textiles 8.6
4,164.8 0.00206
0.21020 Miscellaneous Industrials 22
8 Clothing Footwear 5.2
5,426.5 0.00096
0.09755 Retail 13
9 Wood Wood Products 33.1
5,116.8 0.00647
0.65849 Building Materials 7;
Diversified Industrial 23 10 Paper, Printing Publishing
76.1 15,300.7
0.00497 0.50629
Paper Packaging 12; Media 15 11 Petroleum Coal Products
220.7 10,373.3
0.02128 2.16575
Oil Gas 4 12 Chemicals
169.4 13,139.3
0.01289 1.31239
Chemicals 10 13 Rubber Plastic Products
15.6 6,123.6
0.00255 0.25932
Paper Packaging 12 14 Non-Metallic Mineral Products
76.3 7,608.4
0.01003 1.02083
Diversified Resources 5 15 Basic Metals Products
372.8 18,918.7
0.01971 2.00589
Other Metals 2 16 Fabricated Metal Products
32.9 11,748.1
0.00280 0.28507
Miscellaneous Industrials 22 17 Transport Equipment
27.1 14,393.9
0.00188 0.19165
18 Other Machinery Equipment 23.5
16,014.3 0.00147
0.14938 Engineering 11
19 Miscellaneous Manufacturing 11.6
5,282.5 0.00220
0.22353 Engineering 11; Retail 13
20 Electricity, Gas Water 229.4
25,216.8 0.00910
0.92603 21 Construction
215.0 48,560.3
0.00443 0.45069
Developers Contractors 6 22 Wholesale Trade
216.0 34,033.2
0.00635 0.64606
23 Retail Trade 182.9
34,953.7 0.00523
0.53265 Retail 13
24 Repairs 76.2
13,178.8 0.00578
0.58857 25 Accommodation, Cafes
Restaurants 19.9
16,253.1 0.00122
0.12463 Tourism Leisure 24
←
→
R.W. Faff,
T.J. Brailsford
Journal of
Energy
Finance and
Development
4 1999
69–87
75 Table 1 continued
Petroleum Direct
Coal Australian
Requirement Products
Production Coefficient
Relative ABS Industry Classification
a
AUDm AUDm
DRC
b
DRC
c
ASX Industries ASX Industry number 26 Transport Storage
2,055.0 41,928.8
0.04901 4.98909
Transport 14 27 Communication Services
121.0 15,522.1
0.00780 0.79352
Media 15 28 Finance Insurance
31.9 40,073.5
0.00080 0.08103
Banks Finance 16; Insurance 17 29 Ownership of Dwellings
16.7 46,164.8
0.00036 0.03682
30 Property Business Services 410.3
54,245.6 0.00756
0.76994 Property Trusts 20
31 Government Administration 147.7
37,682.5 0.00392
0.39899 32 Education
0.7 22,631.8
0.00003 0.00315
33 Health Community Services 155.6
33,768.1 0.00461
0.46906 Miscellaneous Services 21
34 Cultural Recreational Services
40.9 13,866.2
0.00295 0.30025
Tourism Leisure 24 35 Personal Other Services
232.8 12,844.3
0.01812 1.84499
Miscellaneous Services 21
a
Australian Bureau of Statistics ABS classification contained in the Australian National Accounts Input-Output Tables.
b
Direct Requirement Coefficient DRC for each industry is calculated as the ratio of Petroleum and Coal Products Australian Pro- duction.
c
Relative DRC for each industry is calculated as the ratio DRCIndustry Average DRC.
76 R.W. Faff, T.J. Brailsford Journal of Energy Finance and Development 4 1999 69–87
ASX 21 industry provides a mixed signal; while it has a large RDRC, this effect will be countered by the positive effect via the inclusion of the Petroleum Products
sub-industry ASX 213. A similar argument applies in the case of the Miscellaneous Industrials ASX 22 because it includes agricultural products, textiles, and fabricated
metal products. Moreover, we are skeptical of any effect revealing itself in relation to the Other Metals industry ASX 2, as it too provides a mixed signal. This industry
has a high RDRC but is also likely to experience an indirect positive sensitivity because oil prices are generally correlated with metals prices.
It is also useful to consider the industries that have a value for the RDRC that is extremely low. One notable case is the ASX 13 Retail industry, which seems to be
consistently low as it relates to ABS 8 Clothing and Footwear, ABS 19 Miscellaneous Manufacturing, and ABS 23 Retail Trade. Another consistent case of low RDRC
is the ASX 11 Engineering industry in relating to ABS 18 Other Machinery and Equipment and ABS 19 Miscellaneous Manufacturing. Accordingly, other things
being equal we predict a negligible role for the oil price factor for these industries. The remaining industries also appear to generally indicate a negligible role for the
oil price factor.
This analysis highlights the importance of conducting the test on industry sectors. The hypothesized differential impact means that aggregate market analysis as pre-
viously conducted in papers such as Jones Kaul, 1996 may mask industry effects. In summary, we predict a positive sensitivity in four oil and oil-related industries
ASX 1, 3, 4, and 5 and a negative sensitivity in four other industries ASX 2, 10, 12, and 14. However, the ability of firms to pass on oil price changes to their customers
and the extent of hedging activity mitigate against finding significant results.
3. Data and research design