g L S )7 C/ oy, Research, and External Affairs WC PAPERS Debt and International Flnance International Economics Department The World Bank August 1 991 WPS 749 Hedging Commodity Price Risks in Papua New Guinea Stijn Claessens and Jonathan Coleman With increasing awareness of commodity price risks and with technical assistance - strategic advice and assistance in institu- tion building and skills training - -developing countries such as Papua New Guinea can learn to use market-based commodity- linked financial instruments to improve their economic manage- ment. The Policy. Research. and Extemal Affairs Complcx distnbics PRE Working Papers todisscminatethe findings of work in progress and to cneouragc the exchange of ideas among lIank staff and all others interested in development issues. These pipems tany the names of the authors, reflect only their views, and should be uscd and cited accordingly. The findings, interprctations, and conclusions are the authors' own. Thcy should not be attributed to the World Bank, its Board of Directors, its management, ur any of iLts member countries. i Policy, Research, and External Affairs Debt and Inte;,iational Finance WPS 749 'Ibijis paper - - a joint product ol tic Debt andi Internaitional Finance and International Trade Divisions, Inteniational Economics Departiicit - is part of a larger clfort in PRE to stu(y' the use of financial ins.ruilcints to mTanagc thc exictrnal exposures ol devcloping couIInties. Copies are ava&lable free from the World Bank, 1818 1I Street NW, Washington, DC 20433. Please contact Sarah Lipscomb, room S7-062, cxtension 33718 (31 palges, wiih figoures and tables). Papua New Guinea faces substantial exposure to export earnings, short-tcrni hledging tools, such price Iluctuaitionis for its major ptiimalr) coiminocd- as options and lutures, could be used effectively. ifv exports: gold, copper, cotiee, Cocoa, logs, Claessens and Colemani design specific financial anlld palmn oil. Its existinig conmmodily uisk strategies that Papua New Guinea could use, and managcnient schemics its ninii-eral slabili/.ation demonstiratc the gains to be made from active futind and agricultural commiillodit) funds -- are risk management. cosily, provide only li iitied protectlion against hlec intpat of fluctuations in comtnodit) prices, Tle lcessons leIarIIe(d IItc not unique. Many aInd arc unablc o l)rovidce protection foI loIIg developing countries are hcavily depen(lent on pcriods, primary cominodities for foreign exchange, and thcir econlomic developmcnt has suffered from (ilaesscris arid Coleman show thlat ma-rkct- thc resultinig risks and instabilities. With in- hased financial instirumienits ate better suiled tO ct-casinig awareness of these risks and with manage exteinal price risk for a countii) that is a technical assistance - strategic advice and pric l taker in wvorld comimiiodity markets. T his is assistanlce in institution building and skills especially itic case forI mineral and cnergy price training - developing countries can learn to use iisks where f'inanicial instrum tcits (such as finanicial instruments to improve their economic comin modity swaps) exist for hedging export management. ea1nin1gs over long periods. For agricultural I'IhC PRE Workinig Paper Setiie dissmniinales the f-indings of uNork under way in the Bank's lPolicy, Rcsearch. and External Al faiis Complex. An ohjeclivc ftiiescries is to ge thetsc findings out quickly, cven ir presentalions are less than fullt pxlished. The finndings, interpretations, and conclusio', in these papers do not necessarily represent OrfiCial Bank policy. Piodu(ced by ihe lPRE Dissemiiination Center Table of Contents 1. Introduction .......... ...........1 2. Background: Importance of External Risk Management to the PNG Economy .1 2.1. Importance of Primary Commodities to the PNG Economy .2 2.2. Nature of Primary Commodity Value Fluctuations .4 23. Problems Created by Commodity and Other Price Instability .6 3. Existing Commodity Risk Management Mechanisms . . .7 3.1. Mineral and Energy Price Risk ..7 3.1.1. Taxation and MRSF .7 3.12. Mineral Resources Development Corporation (MRDC) Rules .9 3.2. Private Corporations ..10 3.3. Agricultural Commodity Boards ..10 4. Impact of Commodity Price Fluctuations on Economic Stability in PNG . .12 4.1. Tax Revenues.12 4.2. MRSF.13 4.3. MRDC and the Private Sector.14 5. Risk Management Schemes and Their Costs and Benefits.15 T.1. Stabilization Funds ................ .......................... 15 5.2. Other Instrument s 16 6. Financial instruments to Manage Risk ..17 6.1. Commodity Futures.17 6 t.2. Commodity Options .18 6.t3. Commodity Swaps .18 7. Mineral and Energy Price Risk Management Strategies .19 8. Agricultural Stabilization Funds.26 9. Conclusions ........ ................................................................... 29 R eferences ................................................................................................................................................................... 30 Appendix . .... 31 Thc authors would like to thank Ron Duncan for his vcry useful comments and World Bank staff in AS5CO for their contributions to this paper. 1. Q IIDf Papua New Guiuea (PNG) faces substantial exposure to price fluctuations of its major primary commodity exports. Existing commodity risk management schemes provide limited protection against the impact of commodity price fluctuations, have high cost, and are not able to provide protection over long time periods. (e.g., the agricultural stabilization funds are effectively exhausted). This paper shows that financial instruments available in developed capital markets are better suited to manage the external risk of PNG than existing schemes, and are less costly. This is especially true for mineral and energy price risks where financial instruments exist for hedging over long maturities. This paper shows how these instruments could be used by PNG. For the agricultural stabilization funds, short-term hedging tools could be used effectively, and, for illustration, a simple hedging strategy is developed for the coffee fund, The paper is organized as follows. In section 2, the importance of primary commodities to PNG is discussed, as well as its exposure to volatile international commodity prices. In section 3, existing commodity risk management mechanisms are discussed, and in section 4, the exposure of economic stability to past commodity price fluctuations is quantified. In section 5, a general overview of risk management schemes and their costs and benefits is presented, and section 6 discusses some specific financial instruments applicable to PNG for external risk management. Risk management strategies for the mineral and energy sectors are developed in section 7, and for the agricultural stabilization funds in section 8. Conclusions of the study are drawn in section 9. 2. Backaround: Importance of External Risk Management to the PNG Econom! This section describes why external risk management is of primary importancc to the PNG economy. First, PNG is highly dependent on the cxports of primary commodities for forcign cxchange carnings, govcrnmcnt revenues, and employmcnt. Second, PNG is a price takcr in the world markcLs of its major primary commodity cxports. During the 1970s and 1980s period thcsc markets have bccn highly volatilc, with 2 large intra-year and inter-year fluctuations in prices, and thus a major source of instability in the PNG economy. Third, PNG's debt structure exposes it to both exchange rate and interest rate risks. Therefore, the use of commodity, interest rate, and currency risk management instruments would be of considerable value by reducing these exposures. 2.1. Motane of Primary CoModities to the PNG Econom The performance of the PNG economy is determined largeiy by the strength of the export sector which is composed mainly of mineral and treecrop exports (see also Table 2.1). These are crucial to the economy in terms of foreign exchange earnings, government revenues, employment, and external debt servicing. However, agricultural production is the major source of employment in PNG. Within this sector, coffee and logs dominate, with the importance of cocoa, copra and cocc.nut oil, and palm oil declining since the mid-1980s. Thbe importance of primary commodities in export earnings is illustrated in Table 2.1. In 1989, almost 70% of the total export earnings of PNG were obtained from the exports of gold and copper compared with less than 50% in 1985. It is projected that the mineral sector will continue to dominate in the early 1990s, contributing a little less than two-thirds of total export earnings in the 1990-1992 period. Outside the mineral sector, logs and coffee contribute the most to export earnings. In 1989, coffee and logs contributed 10.9% and 6.6%, respectively, with cocoa and palm oil at 3% to 4% and copra 1.4%. These proportions are forecast to remain fairly stable in the early 1990s period. The minerals sector makes a large contribution to governmcnt revenues through corporation income taxcs, dividend withholding taxes, and dividends from government equity in mineral prcjects. In addition, there are the import duties and payroll taxes paid by the mining corporations. In total the mineral sector providcd 20% of govcrnment revenues in 1989, and by the end of the decade their share is forecast to rise to over 35%. 3 Tabl 2.1. Contnbuion of Maior Primary PgmModiyEWs to Total Emrt pamin.NG. t. Commodity 1985 1986 1987 1988 1989 1990 1991 1992 Percent Minerals 46S 60.9 61.7 70.6 69.0 6S5 64.1 64.7 Gold 25A 402 413 36.4 25.1 31.2 34.3 40.1 Copper 21.1 20.7 20.2 34.2 43.9 34.3 29.8 24.6 Nonmintemls 535 39.1 383 29A 31.0 345 3S.9 353 Cocoa 75 63 5.3 35 3.8 3.2 2.6 2.6 Coffee 13.4 IS4 16S 9.4 10.9 9.2 7.2 6.9 Copra 5.1 2.0 1.0 1.3 1A 135 IA 1.2 LOP 7.7 5.7 6.9 7.6 6.6 8.9 10.1 9.6 Palm Oil 73 3.9 2.8 1.7 3.1 4.3 5A S.6 Other 12.3 5.8 S.8 5.9 5.2 7.4 12.2 9.4 Source: Based on Table H in Annex Vil of IMP, 1990. Although less important than the mineral sector, the agricultural sector makes a significant contnibution to government revenues though direct taxation of company income, taxes on agricultural exports, and profits from government equity in agricultural projects (e.g, oil palm estates). In addition, there are the indirect taxes imposed on imported agricultural inputs, as weil as the taxes paid by individuals earning agricultural incomes, and the excise taxes levied on items such as fuel, beer, and cigarettes purchased with incomes generated from agriculture. It is estimated, for example that the contribution to total government revenues of the coffee sector alone is as much as 10% (see Brogan and Rewenyi (forthcoming)). The mining sector generates few opportunities for employment in PNG. Most of the capital used in the mines is technologically advanced and is imported. The two major mines, BCL and OTML, together cmploy about only 6,000 people (many of whom are expatriates) which is about 0.3% of thc labor force. The labor force in PNG (estimated to be 1.8 million in 1987) is primarily employed in agriculturc. Within the formal sector, agriculture is the most important source of cmployment, taking about 20% of the labor forcc. Whilc the terms of trade havc movcd against the agricultural exports of PNG sincc the mid-1960s, rcal 4 consumption and investment have been maintained by overseas borrowing. As a result, cxternal debt has increased dramatically from less than $200 per capita in 1970 to more than $1,200 currently. The debt service to exports ratio is currently at about 30%. In the 1988 government budget, and interest paymtents alone made up almost 7% of total expenditures. External debt (servicing) has further been influenced by two external factors: intermational interest rates and cross-currency exchange rates. The influence of movements in international interest rates on PNG's debt service obligations has been relatively small compared to many other developing countries, since a considerable part of PNG"s long-term debt is of a fixed rate nature (approximately 45 percent). Still, a change of one percentage point in the interest rate alters debt service by about $13 miUion. The influence of exchange rate movements on the level of ntebt measured in US dollars has been large since a significant part of PNG's debt is in non-dollar currencies (approximately 60 percent). Over the period 1985 to 1989 the absolute value of the currency valuation effect on debt stock has, on average, been about $72 million annually or about 3 percent of the debt stock (see also the Appendix). 2.2 Nature of Primaa Commodity _alue Eluctation The instabilily of export revenues of primary commodities is associated with fluctuations in both quantities produced and prices. Indexes of the value, volume and unit value of the major primary commoditics of PNG between 1985 and 1989 and projections for 1990 to 1992 are reported in Table 2.2. Also reported in Table 2.2 are the coefficients of variation (CV) (the ratio of standard deviation to the mean) which provide a crude measure of instability. The CV for the index of the value of mineral exports for the 1985-1992 period is almost 20%, indicating that cxport value is quitc unstable. The value of copper exports is espccially variable with a CV of 36.5%, while the CV of gold was 19.0%. Also, copper prices werc highly unstable, recording a CV of 24.4%. The index of the value of nonmincral exports was more stable than the mineral index, with a CV of 14.4%. The most important commodities--logs and coffcc--reported CVs of 20.6% and 34.5%, rcspectively, which arc both lower S than the CVs for the other major commodity exports, except for gold. Ta le 2.2 P-n Mew -uinea: -zMEW Value. Volume. and Unait Value, b Maie CommOdity. 19951992 Commodly 1985 1986 1987 1998 1989 1990" 199177 199277 C$'1 YAks Minerals 68 9S 86 131 128 100 96 111 19.3 Gold 78 132 122 141 98 100 108 14S 19.0 Copper S9 62 54 121 156 100 8S so 36.5 Nonminerals 152 124 109 109 116 100 96 107 14.4 Cocoa 229 20S 1SS 134 148 100 81 94 345 Coffee 139 169 164 123 144 100 77 84 26.3 Copra 327 133 67 104 117 100 91 91 59.9 Logs 83 65 71 116 90 100 111 121 20.6 Palm Oil 167 93 S8 47 88 100 122 143 37.2 Volume Minerals 81 11S 104 136 118 100 108 129 14S Gold 72 125 115 124 102 100 110 149 188 COpper 90 1OS 94 146 133 100 106 110 16.4 Nonminerals 82 83 86 91 92 100 100 106 8.9 cocoa 83 82 80 78 115 100 92 103 111 Coffee 68 64 76 87 87 100 87 90 13.7 Copra 135 IS0 131 117 104 100 88 89 17.9 LOP 87 90 93 114 95 100 112 lS 11.0 Palm Oil 70 78 72 48 76 100 118 129 29.3 Uoint Value Minerals 84 83 83 96 108 100 89 86 95 Gold 108 106 106 114 96 100 98 97 5.6 Copper 66 59 57 83 117 100 80 73 24.4 Nonminerals 185 149 127 120 126 100 96 101 22.3 cocoa 276 250 194 172 129 100 88 91 39.9 Coffee 204 264 216 141 166 100 89 93 38.1 Copra 242 89 51 89 113 100 103 103 47.7 Lop 95 72 76 102 95 100 99 103 11.9 Palm Oil 239 119 81 98 116 100 103 111 38.0 Source: Based on Table iii in Annex VII of IMF, 1990. I/ Projections. 2/ Coctficient of Variation. With thc cxccpzion of gold and logs, thc unit value variability of each export commodity in the table 6 is greater than its production variability. This reflects the high degree of instability of intcrnational agricultural commodities prices. Also of interest is the fact that the commodity unit values have declined over the period, especially for coffeeand cocoa. The dependence of PNG on primary commodities will be strengthened with expected developments in the energy sector. Oil export earnings could amount to as much as one-third of mineral export earnings by the year 2000, equivalent to 25% of total export earnings. With greater dependence on oil cxports the economy of PNG will open itself to risks associated with fluctuating oil prices. Recent events have been a reminder that oil prices are highly unpredictable which suggests that risk management instruments to lower these risk will be of great importa-ice in the future . 2.3 Probems Created by Cmm-odity and Other Price Instability The previous section clearly demonstrates that commodity price, exchange rate and interest rate instability has had and will continue to have major impacts on PNG's macroeconom; through its cash-flow effects on export earnings and the relative burden of debt service. The impact of volatile external prices does not, however, limit itself to its contemporaneous effect on cash-flows but also impacts on production and investment decisions. For PNG, the problems involved in developing the non-mining sector of the economy can, in part, be traced back to the large dependence on mineral exports and the volatility in mineral prices. In times of high mineral prices and high real exchange rates, the international competitiveness of the non-mining sector deteriorates and little investment occurs. However, in periods when mineral prices and the real exchange rate fall, the non-mining sector may still not develop, as investors realize the situation can casily revcrse itself in the future, rendering investments in the non-mining sector possibly unprofitable. Ilistorically, the CV of oil prices has bccn hctwcen 20% and 30%. 7 3. Exslin CommadiAisk Mananemet Mechaisms In this section the linkages between the commodity sectors and other sectors of the economy are explored in greater detail and a description of the important institutions and mechanisms through which commodity prices affect the economy is provided. There are three main entities which bear the risks associated with fluctuating commodity prices. These are (i) the Government of PNG (GOPNG); (U) the private corporations that operate in these sectors; and (iii) the agricultural marketing boards for coffee, cocoa, copra, and palm oiL We discuss the allocation of mineral and energy price risks first, followed by agricultural price risks, and the specific risk management schemes in place in each sector. 3.1. Minera and E Price Risk The GOPNG is affected by price changes in tdie form of tax revenues channeled through the Mineral Resource Stabilization Fund (MRSF), and through the, Mineral Resource Development Corporation (MRDC) (through its equity stak. n mines and energy projects and its responsibility for raising the necessary funds to rinance the equity participation in new projects). 3.1.1. Taxation and The structure of mineral resource taxation in PNG reflects the authorities' objective of providing adequate incentives to producers while ensuring that the government is able to secure most of any windfall profits (see further Coopers and Lybrand (1989)). The tax regime in place for both the mining and the pctrolcum projects places a heavy tax burden on the more profitable operations while minimizing tax requirements from marginal projects. Mining cnterprises arc subject to an effectivc tax rate of 46%. Highly profitable operations are subject to an additional profits tax (APT), which can result in marginal rates of a;;nost 65%. The mining APT is payable when the project shows a return on investmcnt abovc a specificed rate. As a result of the APT, the government's revenues increascs sharply in periods of high commodity prices. 8 Petroleum projects are subject a company tax rate of 5;9%. The APT for the petrel' -um .ndustry is payable at a rate of 50% once the project has achieved a 27% nominal rate of return after income tax. The Mineral Resource Stabilization Fund (MRSF) was established by an A.., of Parliament in 1974, with the objective of reducing the impact of fluctuations in mineral revenues or. the government budget. Under this legislation, all dividends in state shareholdings, companv income taxes, and dividend withholding taxes (identified above) from all designated mining operations must be paid into the MRSF. The assets of the MRSF are held by the Central Banik of PNG (BPNG) which invests them in securities abroad, primarily in the form of interest-bearing deposits and central bank securities. The surplus funds, managed BPNG, constitute the main source of reserves. Drawdowns from the MRSF are determined on i;- basis of recommendations submitted to the Board of Management of the MRSP. The general rule for withdrawal is that the amount to be drawn down should ensure that the Fund is sustainable in terms of real purchasing power over the next five years (see also Guest (1987)). The Board of the MRSF is bound to make forecasts of future receipts for eight years ahead and (implicitly) for inflation forecasts for five years ahead. The commodity prices implicit in tke forecasts should not vary more than 10 percent from the historical moving-average of commodity prices (20 years for copper prices, the preceding year for gold and silver). In practice there has been considerable flexibility in the operation of the MRSF and a revision made to the MRSF Act in 1987 allows the government greater discretion than before. This, while providing greater flexibility in the use of mineral revenues, carries the risk of larger increases being allowed in drawdowns and expenditures in anticipation of future growth in these revenues. The contributions to and drawdowns from the MRSF have varied substantially over the past decadc. The degree of stabilization the MRSF has provided can be quantified by comparing the Cv of its outflows over the pcriod 1980-1988 (45%) with the CV of revcnues (56%). This suggcsts that very little stabilization of the government's budget has taken place by placing MRSF 9 between tax receipts and inflows in the budget: only a relative reduction of about 20%2. The result is to be expecced given the implicit use of the moving avera2-e price with short time periods (one year for gold and silver) as an indicator of the future price and the (recent) flexibility in the rules. 3.1X Min:rp1Kesources l)pnentCorporation (MRDC) Ruiles The govermment has followed a policy of taking an equity share in all major mineral projects. Although these investments have been characterized by a high degree of risk, there has been popular support for the principle that the government should maintain a share in the ownership and control of projects involved in exploiting nonrenewable resources. The equity-participation in new mineral projects has taken place lusing the MRDC as a vehicle. In the case of mining projects, the govermment has reserved the right to take an equity share ok tip to 30%, although in most projects to date it has taken a 20% share. The expectation is that, in some form or another, the state, provinces and landowners will, through MRDC, retain an equity stake in the mining projects. For petroleum projects, the government reserves the right to take d 22.5% carried interest. MRDC has financed its equity participation in the different projects in a variety of ways from deferred payback on future dividends--in which case the foreign investors effectively provides the financing--to loans obtained through external commercial borrowing (with a government guarantee)3. More re ently, MRDC has relied on foreign financing. MRDC receives their dividends on its equity stake and pays any excess of dividends over financing costs to the r[FP. Through its participation, MRDC is exposed to commodity price risk since its expenses (interest and principal payments on foreign loans) are not dependent on commodity prices whereas its rcvcnucs are. The substantial level of government equity in new mines ^nvisaged over the next few years will This result is conflrmcd by the analysis of Guest, which dcrivcs thc similar result that the NIRSI has reduced the instability of mineral revenues by only 30 percent. 31'frcctively, the Departmcnt of Financc and Planning (DI') has arrangcd the financing under its nanc and thcn passcd it on to %IRDC. 10 require large amounts of new financing, (commercial external borrowings) which arc estimatcd to amount to about $500 million in 1991 and 1992. This will greatly increase MRDC's exposure. 3.2. Prinate Corrations Foreign companies operating mines in PNG are exposed to commodity price risks through thc impact of price changes on the cash-flow derived from the mines (dividend remittances and other transler payments to the parent company). Most foreign firms have put in place some risk management program to protect cash- flow streams to parents against commodity price swings. It a-ioears, however, that the risk manage:mnit is dlone at the parent, off-shore level, i.e. net profit remittances received by the parent from the subsidiary are hcdlged at the off-shore level, leaving export receipt and taxes exposed. This implies that much of the commodity price exposure remains at the subsidiary level and consequently that the PNG economy remains ecposed to fluctuations in the price of its main exports. 33. Abr,icultural Commodity Boards Given the importance of agricultural commodity exports to the economy and the volatility of international prices, stabilization schemes have been established in PNG since the 1940s for the impor.iant export crops in Papua New Guinea (PNG) -- namely cocoa, coffee, copra and palm oil. The four schlciiems zirc similar in design (see Figure 1). A threshold pr;ie is determined equal to a ten-year moving averate 1 F(F)l prices, adjusted for inflation. Then a buffer zone is set at 5% above and below the thrcshold price in wvhich no bounties or levies apply. When the current FOB price is more than 5% above the threshold price, levies are imposed on producers at 50% of the difference between the threshold price and the current FOB pricc. The levy revenues are paid into a commodity stabilization fund. When the current FOB price is more than 5"s, below the threshold price, bounties are paid to growers at 50% of the difference between thc curreint I01)B prrice and the threshold price. Bounties are paid out of the stabilization fund. 11 price/ton/ Farner Figure 1 ~ ~ ~ ~ ~ ~ ~ ~ the Unt_ recently,_-he9schemes5were succeSfulinstabilizing of thresoold prices..w.i.dn ......'..-' 4 jj / ~~~~leuyg A obUt uFfer / 1 3 ~~~~~bounty m// no levy zone Figure I Until recently, the schemes were successful in stabilizing producer prices. Howcvcr, in ilic midl- aInd late-1980s, commodity prices fell in real terms, so that current FOB priccs were consistently below thc tlhrcslold price. As a result, subsidy payments were made to growers over an extended period. This lcd to thc eventual exhaustion of the cocoa, copra and palm oil funds, and the coffee fund is expected to run out carly in 1991. To prolong price support, the funds were kept solvent through government financial contributions, commiiier-ciall bank loans and STABEX transfers. In response to these problems, in November 1989 the government decided to implement it number of interim measures which would provide support to the agricultural export sectors while a new approach to pricc stabilization was found. The objective of these new measures is to give price protection to growers while thiv adjust to lower international prices without the support of bounty payments from the stabilization funds. Instead of establishing a threshold price based on a long-run moving average, support durinig the atlisltlllct pcriod is based on the difference between the estimated costs of production and the international price. T'h. level of price support will decline over an adjustment period of about three to five years dcpendiing on ihc commodity. By the end of this period producers will face international priccs. The loans used to finanec pricc support payments are to be repaid by the commodity boards from cxport rcvenues when the international pl ice exceeds the support price. 12 The rationale for this policy is that without such a scheme agricultural incomes would decline substantially with a significant proportion of the estatc sector going out of business. With support, as well as initiatives to improve productivity, such as extension and research, the sectors will be able to adjust to low prices and to gain international competitiveness, as well as being in a position to repay existing loans. 4. Impact of Commodity Price Fluctuations on Economic Stability in PN(G In this section we provide empirical estimates of the magnitude by which each institution is affccted by commodity price risk. We identify the following parties: tax revenues, MRSF, MRDC, and the private sector. 4.1 Tax Revenues Historically, PNG's total tax revenues have been very sensitive to variations in commodity prices. This can be estimated by running the regression: (1) t a + ,Pt1 + -YPt2 + ... + error where TRt are percent changes in tax revenues in period t, and Pt, are percent changes in the prices of relevant commodities (e.g., copper and gold)4. From these regression cquations, the clasticity of tax revenues with respect to the copper price (both expressed in percentage annual changes of dollar values) over the period 1976- 1988 was about 0.25 (with a t-statistic of 3.3 and a R2 of 0.56). The clasticity with respect to gold prices was about 0.18 over the same period, but not significant. A similar regrcssion was done for export earnings. Thc elasticity of export earnings with respect to copper prices was about (1.18 (with a t-statistic of 1.37 and a R2 (if "Scc further G(emmill (1985) and Kolb (1985) on how to cslimnatc thcsc sensitIvifics. 13 0.135). The elasticity with respect to gold prices was about 0.738 over the same period (t=2.84, R2 = .45)5. These results suggest that the exposure of tax revenues to export prices is quite different from the relationship between export earnings and export prices in the case of the gold price, but similar in the case of the copper price. Since the average share of copper and gold in export earnings over the 1976-1988 period was about 25% and 30% respectively, these elasticities indicate also that movements in volumes cxported have partly offset the effects of price movements in the case of copper exports and exacerbated those movements in the case of gold exports. 42 MRSF Similar regressions were performed for the dividend stream on the government's share in the mining projects accruing to MRSF. The elasticity for gold was 1.22 (R2 = 0.28, t = 2.06) and for copper 0.72 (R2 = 0.05, t= 0.77). The regression coefficients for the annual levels of dolar dividends (in millions) on the level of prices for the period 1981-1988 were 0.076 for gold (RG = .43, t=2.14) and 0.005631 for copper (R2 = .42, t=2.11). Since the last regression is in levels, the coefficients measure the exposure of the dividend stream to prices and can be interpreted as the quantity of physical commodities 'received' by the government each year. These quantities are equal to 2.56 tons of gold (converting ounces to tons) and 5,631 tons of copper (or 8 percent of gold exports and 2.5 percent of copper exports). Regarding future price sensitivities, the World Bank (1989) reports results for two scenarios: in scenario I the price of gold is 15% lower than in the base case and the price of oil stagnates in rcal terms; in scenario 2 the price of gold rises proportional to international inflation. The diffecrcnce in thc currcnt account between thcse two scenarios is 7.5 percent of GDP and in the fiscal balance is 6.7 perccnt of GD?, indicating thc largc scnsitivity of both aggregates to international commodity pricc movements. ' Ovcr a longcr period the elasticity of cxport earnings with respect to copper priecs is 0.54 (with a t-statistic of 2.97), and the clasticitv with respect to gold prices is 0.64 (with a t-statistic of 2.11). 14 43 MlDQCadlhe Private ScLtor Sensitivity scenaros regarding future prices can also be performed on the profitability and resulting tax revenues in the case of an individual mining operation. This was done for a mine similar to the recently opened Porgera mine, which largely produces gold. Based on production estimates, costs of production, and current tax regulations profits, tax receipts and dividends paid abroad (the excess cash-flow after subtracting the government's share) are calcuated under different gold price assumptions. Present values of Dividends and Taxes £00 w~~~~~~~o 700- a m0 SOO - 400- 70) '000 0- S300 $325 S310 £375 £400 £425 £41O05475 %S00 1550 5600 Gold b-Ic.. S/ounc ~Figure 2 Figure 2 plots the results for the present value of dividends paid abroad and the present value of gross receipts to MRSF (tax receipts, royalties and dividends) over the life time of the project. As one can observe, the prcsent value of MRSF receipts, PVT, is more sensitive with respect to the gold price than the prescnt value of dividends paid abroad, PVD. As the gold price increases, the APT comes into cffcct and raises the PVT rclatively morc than the PVD. In rcgrcssing the PVT on the gold price, the slope is around two, while thc slope of a regression of the PVD on the gold pricc is only 0.75. Tlhus, the govcrnment is rclatively (about thrcc times) morc cxposed to gold price risks than the forcign investors arc, since ilicy share less in the upside of 15 a price increase while the government shares more because of the APT6. 5. Risk Management Schemes and Their Costs and Benefits The above section indicated the need for risk management, especially commodity risk management. There are many different forms of commodity risk management that can be used by developing countries. We will discuss here the ones in use by PNG. 5.1 Stabilization Funds The principle behind the agricultural product stabilization funds and MRSF in PNG is that in periods of decline in commodity-related tax revenues the government draws on its reserves to finance its normal expenditures. The experience with commodity stabilization funds in other countries has shown, however, that these funds are seldom sufficient to sustain expenditures in times of prolonged downturns in commodity prices and that consequently the funds only insure against short-lived and temporary declines in commodity prices (Gilbert (1990)). Evidence of this is provided by the agricultural stabilization schemes which are effectively depleted. The PNG agricultural stabilization schemes were designed originally to stabilize price around the long- run average price. The non-stationarity of commodity prices means that use of the moving-average of past prices is unreliable as an estimate of the long-run price, and that the scheme is likely to fail. As reported above, the agricultural stabilization funds in PNG have indeed become exhausted following the persistently low prices of cocoa, coffee, copra and palm oil during the late 1980s. Thcre are many similar cxamples from other countries (e.g., coffee and cocoa stabilization funds in Cote d'lvoirc and Cameroon). MRSF is the onlv stabilization fund in PNG that has so far performcd successfully for a prolonged period. The recent closure of BCL has shown that cvcn the MRSF can only provide vcry limited insurance against a major shortfall in 6In terns of Icvcls. the cocfficicnt for PVD is about $80,000 for the govcrnmcnt's sharc. 16 government revenues and that additional support for PNG's balance of payments has becn necessary. Adjusting the rules of a scheme will not prevent the funds from exhaustion over the long-run, unless the effectiveness of the scheme in stabilizing prices is drastically reduced. For example, using historical data it can be shown that if the moving average period were to be shortened and the inflation adjustment were eliminated from the PNG agricultural schemes, the funds would not now be exhausted. However, while such changes would certainly reduce the frequency of depletion of the stabilization funds, the trade-off is that the price paid to producers would be less stabilized. Another modification to the design of the schemes which would slow the exhaustion of the funds would be to allow only 50% of the fund be paid out as bounties in any one year. In this case the fund would never become exhausted (although it could asymptotically decline to zero), but it would become progressively weaker in its ability to stabilize producer prices. The agricultural schemes in PNG have failed largely because of the difficulty of determining the long- term price which in turn determines the level of withdrawals. Funds in other countries have failed for this reason but also because of lack of discipline in accumulating reserves in times of high commodity price. The increased flexibility with respect to drawdowns from the MRSF thus raises some concerns with respect to budgetary discipline. 5.2 Other Instruments PNG has also benefitted of the international commodity agreements for cocoa and coffee. Most of the international price stabilization schemes have collapsed in recent years because of a breakdown in cooperation among its members or because of the exhaustion of support funds. PNG has also drawn funds from compcnsatory financing schemes (such as STABEX/SYSMIN facilities, and CCFF), which can serve as risk managemeni schemes. However, they have the drawback of their limited and uncertain availability, as was shown following the reccnt closure of the Bougainville copper and gold mine (BCL). 17 6 Financial Instruments to Manage External Risk The failure of the international commodity agreements to stabilize prices on a reliable basis and the possible drawbacks of compensatory agreements and domestic stabilization schemes point to the importancc of financial market instruments for commodity price risk management purposes. The most important financial instruments from PNG's point of view are commodity futures, commodity options, commodity swaps and commodity-linked fnance (loans and bonds) (Masuoka (1990)). More specifically, forward contracts, futures contracts and options could be used to eliminate price exposure over a short time period (e.g., one year) for the agricultural commodities. Swaps could be used in the mineral and energy sector for longer horizons. The nature and usage of these short- and long-term hedging instruments in the context of PNG is now further described. 6.1 Commodity Futures As an example, the Cocoa Marketing Board in PNG may wish to set a guaranteed price to producers for the coming season without incurring significant financial losses if international prices change suddenly. Based on historical export patterns, the Cocoa Board could predict fairly accurately the quantities of cocoa available for export during different months throughout the year. Then it could sell futures contracts at the beginning of the season for each of the delivery months in the coming year in proportion to the volume available for export in each of those months. The guaranteed price at the beginning of the season to producers could be set as the weighted average of the futures prices (with weights given by the export volume). As exports are sold at international prices throughout the year, near-by futures contracts (i.e., those closest to expiration) could be purchased (thus offsetting the original short position) at prices close to the international price. Such a strategy would eliminate intra-year price risk for the Marketing Board. 18 6.2 Commodity Options Commodity options (on futures contracts) could also be used by the commodity boards. For example, say the Coffee Marketing Board wished to hedge the future price of its coffee sales with options instead of using futures or forward contracts. Suppose the Board wished to scll 50,000 tons of coffee in six months timne and wants to receivc a price of at least $2,000 per ton. In this situation, the Board purchases put optiorns giving it the right to sell coffee futures contracts at a price of $2,000 per ton. Say the premium quoted is $50 per ton so that $2.5 million is p2id to cover the entire 50,000 tons. If after 6 months the price is greater than $2,000 per ton, say at $2,200 per ton, the Board would not exercise the options and receives $110 million in revenues. If, however, the price falls to $1,800 per ton, the Board will exercise its options, enabling it to sell futures contracts at $2,000 per ton. The futures contracts can then be bought back at $1,800 (since the spot and futures prices are always equal at the expiration of the contract) making a profit of $200 per ton. The Board then sells the coffee at the spot price of $1,800 per ton. Combining the $200 per ton gain on the futures contracts and $1,800 per ton on the physical commodity gives an overall price of $2,000 per ton and total revenues of $100 million. The coffee options provide price insurance to the Board, guaranteeing it at least $100 million in revenues in exchange for an insurance premium of $2.5 million. 6.3 Commodity Swaps Assume that PNG and a big German company agree upon a long-term export contract in which the German company buys 2 million pounds of copper every six months over the next five years and pays the then current copper price at each six-month interval. Assume also that the PNG exporter wants to "lock-in" the dollar value of thcse revenues now. The exporter now enters into a commodity swap contract with a U.S. bank. Assumc that the term for the swap is $1/lb. (indicative). The U.S. bank agrecs to pay the exporter U.S. $2 million cvcry six months for the next 5 years. The exportcr agrecs to pay thc value at the spot price of 2 million pounds of copper on the same dates when the bank is duie to make its payment--in effect a "diffcrcnce" 19 check settles the transaction each 6 months. Thus, the commodity swap contract is, in cffect, a series of 10 forwari contracts lined up over the next 5 years and has locked the price at $1/lb. PNG has sufficient access to private financial markets to be able to use any of these financ al risk management techniques. Given the increase in exposure of the PNG economy to commodity price risk, the economy will increasingly depend on a handful of commodity exports for its growth and development. Therefore, commodity risk management should become an integral part of the GOPNG's economic strategy. We discuss in the next two sections some possible ftnancial risk management techniques for the mineral and energy and agricultural stabilization funds respectively. However, PNG will need to establish first a proper institutional framework and overall strategy and acquire training. Only then specific risk management operations can be implemented. Since PNG is not only exposed to commodity price risks, but also to other forms of external risks arising from its financial liabilities and assets (i.e., exchange rates and interest rates), it is furthermore important that the links between different exposures are taken into account. The interactions between the different forms of external risks which need to be taken into account are the following. First, movements in cross-currency exchanges rates may offset (or exacetbate) movements in primary commodity prices, implying that both commodity and exchange rate management may need to be modified. Second, there can be an inverse relationship between commodity prices and quantities traded. Depending on the elasticity of supply and demand, the effect of price changes on export revenues or import expenses can be offset by changes in the quantity of goods exported or imported. This may reduce the need for hedging price risks as a way to hedge revenue or expensc risks (Coleman and Qian (1991)). 7. Mineral and Encrgy Price Risk Managcmcnt Stratcgics As discussed above, the MRDC docs not insulate thc governmcnt budget from Ilic risks associatcd with commodity priccs because MRDC's liabilities (loans) and assets (dividend streams) arc not matched. This is 20 because its obligations on exdermal loans (debt service payments) are independcnt of commodity prices while its dividends are highly dependent on commodity prices. A matching between assets and liabilities over the long-term can be made with either a commodity swap or commodity-linked finance. An example is when MRDC has an equity stake in a copper mine. In that case, its revenues are sensitive to the copper price and a copper swap can be used to convert copper-price sensitive cash-flows into a certain cash-flow stream which can be used to service obligations on conventional loans. As mentioned by Masuoka (1990), this structure was successfuUly used in the case of the Mexicana de Cobre copper mining company. The notional amount of copper swap would depend on the sensitivity of the dividend stream to the copper price. Based on our earlier analysis, the annual value of dividends (and thus MRDC's revenues) can be expressed as: (2) Dt = a + ,Pt where D is the value of dividends and P is the copper price. Graphically this is shown in figure 3. In section 4, B was estimated to be about 5,631 tons. This implies that for every one dollar change in the price per ton of copper, MRDC's annual revenues change by approximately $5,631. This is equivalent to MRDC receiving (having a long position of 5,631 tons of copper annually, since the value of owning 5,631 tons of copper would change by an equal amount as a result of a price change. This long position of the MRDC in copper can now be hedged by a commodity swap. The swap would oblige PNG to pay annually the spot value of 5,631 tons (P*$P1, which is about $15 million dollars at the current spot copper price) in cxchange for a fixed payment received. The tixed payment would depJ,nd on conditions in the market for copper-swaps and on the futures prices for copper, but for iltistrative purposes a Fixed price of $2,500/ton can he uscd. The coppcr-price sensitive dividend stream will tlien be matched with 21 the copper-price sensitive obligation on the swap. The net result will be a cash-flow stream which is (1argely) independent of copper price swings and °t which would thus, without any risk, be available to service conventional loans. Commodity-linked loans can achieve the same result. Consider the case of the participation of MRDC in a gold-mine, (e.g. a mine similar to I n,^: t Porgera). Once the project comes on stream, the Figure 3 dividend payments received by MRDC will depend on the price of gold. It was shown above that for every dollar change in the price of gold, the present value of a dividend stream to MRDC coming from a project like Porgera changes by approximately $80,000, or, equivalently, the annual dividend stream changes by about $8,ooo7. In order to hedge this risk, the financing of the equity-participation by MRDC should consist of an obligation whose servicing also changes by $8,000 for each dollar change in the price of gold and for the remainder of a conventional loan. A gold-loan could constitute the price sensitive part. Gold-loans stipulate payments in terms of ounces of gold and the costs of servicing the gold loan would thus vary one-to-one with the price of gold. In every period, MRDC would want to owe 8,000 ounces of gold, since this obligation would vary in the same way as the dividend stream with gold price changes. Currently, gold interest rates (gold fees) are between 2% and 3%, so that MRDC could borrow betwecn 250,000 and 400,000 ounces of gold. At currcnt gold prices, this means MRDC could obtain betwecn $100 million and $150 million dollars in a gold loan. The rcmainder would havc to be borrowed in a 'Notice that we measure here thc sensitivity of thc present value of dividend strcams and not thc annu;al dividend streams. Thc prcsent value of all future dividcnds will be more sensitive to assumptions about thc futurc the gold pricc than the annual sircams. Approximately, thc diffcrcnce will be the factor 1/(discount rate). 22 conventional dollar loans. The combination of the gold-denominated and the vonvcntional loans would perfectly hedge MRDC. Similar rinancial hedging tcchniques are possible for MRSF. The current investment policy for the reserves of the MRSF is that they are invested by the Ccntral Bank in rclatively low-yielding, safe assets. This policy does not seem to provide PNG with the best mix of return maximization and risk minimization. Even though using foreign exchange reserves provides a smoothing mechanism, it does not involve laying off risks to other parties and is an expensive self-insurance scheme, because the fund's asset returns are unrelated to commodity price movements and because it cannot sustain a prolonged decline in commodity prices. Further, the fund tes up a significant amount of foreign reserves. A better reserve management policy would be to manage reserves through commodity loans, commodity swaps, or short-term commodity hedging tools. Commodity swaps are the most suitable. Based on projected revenucs streams that depend on the price of a particular commodity, MRSF could enter a commodity swap with a foreign frinancial institution (altogether MRSF should eventually enter several swaps, i.e. separate ones for gold, copper, and oil). As an example, consider MRSF's gold-dependent revenues. MRSF received K42.5 million in total revenues in 1988. Based on the production numbers of BCL and on prevailing gold and copper prices, it is estimated that revenues related to gold-mining amounted to about 30 % of total revenues or 12.4 million Kina (or $14 million) in 1988. This revenue stream depends on the gold price and is equivalent to MRSF holding a long position in physical quantities of gold. Using the 1988 gold price, this long position in tax revenues was equal to about 33,300 ounces of gold. Thus, for every dollar increase in gold price, annual tax revenues of the 89lowever, since thc gold loan is likely to bc in exccss of MRIC's financing necds. thc rcmnaindcr could hc invested in convcntional sccuritics. 23 MRSF change by about $33,000, or the change in the value of 33,000 ounces of gold9. 6MRSP could now enter a gold swap with a foreigm financial intermediary with an interest payment of 33,000 ounces of gold (based on a spot price of 425 S/ounce with a notional interest payment of $14 million) in which it would effectively sell gold at a fixed price for, say, the next 10 years. The ngjQnnlprincipal amount would depend on the current market gold fee; at a rate of 2% it would amount to about 1,650,000 ounces or about $700 million10. In this way, MRSF pays the third party at certain times the equivalent value of 33,000 ounces of gold at the then prevailing spot price--which exactly offsets the tax and dividend receipts it gets from the gold producer. In exchange, it receives from the third earty (commercial bank) a fixed payment of $X times 33,000 at each date (where X depends on the gold fee, the current spot price of gold and the current interest rate), and MRSFs tax revenues for the next 10 years are effectively fixed at $X times 33,000. Swap transactions will not result in a perfect hedge since MRSF revenues do not depend linearly on the underlying price of the commodity, but the revenues depend in a non-linear way on commodity prices and go up progressively when commodity prices rise. Considering the gold sector only, the dependence of the revenues paid into MRSF can be represented as:11 (3) Rt = a + pQ9max[Pt-F-O + AQ*max[Pt-M,O0 where Rt is the revenue going to MRSF, P is the effective corporate tax (the combination of the regular 91he carlier analysis of the gold mine indicated that the regression coefficient of the net present value of taxes (NPT) on the gold price was about 2. This result is confirmned in a regression of the tax revenues paid into MRSP on gold prices which has a slope of about 0.23. Sincc N'P1 is the discounted value of all future taxes, it will be more sensitive to the gold price by approximattly the factor 1/(discount rate), or about 10. '0Notc that the notional principal amount is never cxchanged between the two parties. l.ffectively only intcrcst payment are settled on a nctting out basis. 1tTbis example concerns the revenucs after depreciati3n allowances have expired and the procct gencrates a positive profit. 24 corporate tax rate (currently 35%) and the withholding tax (17%), making for an effective tax of 46%), 0 is the quantity exported, F is the fixed costs per unit for the producer (including depreciation allowances in the early years of the project), P1 is the gold price, and A is the APT tax rate which comes into effect when profits exceed a benchmark, which is assumed to occur when prices P, exceed the level M. This exposure of revenues to gold prices can be hedged using swap transactions (as outlined above) combined with long-dated options. Once the company has largely depreciated its fixed costs, i.e. F = 0, the swap transaction would involve a (notional) amount of gold equal to P*Q which would hedge revenues on account of the regular profit part. This would result in more stable streams to the general budget. In addition to the swap, MRSF could sell today a series of long-dated call-options to a third party with exercise prices of M, for amounts equal to AQ and with maturities on which tax receipts are due in future years. The sale of the call options would give PNG a premium income today. This premium could be invested in safe foreign obligations (e.g. commercial bank deposits or Treasury bills or bonds), from which an annuity could be passed on to the general government budget. The gold swap would now hedge the normal corporate and dividend withholding tax revenues, and paymnents on the swap will be matched with tax receipts at each maturity date. The options would hedge the APT. At the maturity date of the option, the commodity price could either be below the level M (at which the APT is not in cffect) or above it. When it is below, the call is also out-of-the-money and MRSF would not be required to make payments to the third party. When the call is in the money, the paymncnt on the call will cxactly be offset by the tax revenue to the MIPSF from the APT. Again, the transaction would result in converting a risky tax revenuc income stream into a ccrtain yield on the invested premium. These transactions arc shown in Figurc 4. It is clear that the nct payoff thcre is almost indcpcndent of the price of gold as it is horizontal for almost all price ranges. The transactions result in MRSF largelv locking in its revenics over a long period independent of thc gold price. It also allows the MRSF to pass on a fixed stream of payments to the general budget, and not to 25 Hedging Taxes Using Swap and Option ProfIt at $400, APT at $650 250 -^ 200 f 100 - 0 0s 0 A -50 -100 300.00 400.00 500.00 600.00 700.00 800.00 Cold Price a Taxes + Swap 0 Option A Net Payoff Jrigure 4 rely on formulas based on expected prices or ad-hoc decision making. The use of short-term commodity hedging instruments would be similar to that of swaps. An example using commodity futures to hedge carrent year tax revenues against copper price risks would be the following. At the beginning of the fiscal year, PNG would sell copper futures contracts with maturities spread out over the year. The amount to be sold would depend on the sensitivity of the tax revenues to copper prices. The carlier analysis showed that this elasticity was about 0.25, implying that PNG would have to sell futures with a contractual face value cqual to about 25% of anticipated tax rcvenucs. This would cffcctivcly reduce as much as possible the combined effcct of price and quantity unccrtainty on next pcriod tax rcvenue unccrtainty. 26 8. AgiQltural Sabilionu In the description of the fiancial instruments by Masuoka (1990) it was indicated that for most agricultural commodities the longer term hedping instruments were not yet well developed. For these commodities, however, the short term hedging instruments are in general easily available. Since the four agricultural stabilization funds have very similar main features, we will develop below a hedging strategy for the Coffee Marketing Board using coffee futures traded in New York. While the prime objective of the coffee fund is to reduce the effects of sharply changing prices over time, substantial problems can be associated with sudden price movements within the year. This was illustrated very recently with the breakdown in the International Coffee Agreement, leading to coffee prices declining from $3,356 per ton in January 1989 to $1,515 per ton in December. The largest month!y decline was between June and July, with price falling from $2,762 per ton to $1,942 per ton, a reduction of 30%. Intra-year price risks can be hedged using futures contracts. A possible hedging strategy would work as follows. The Coffee Marketing Board offers producers a price for coffee which is fixed for the entire year. This price is set at the beginning of the year and is equal to the weighted sum of the prices of coffee futures contracts which will mature at various months throughout the coming year. The weights are determined by the quantities of coffee available for export in the months between each contract expiration. For example, say in December 1988 the Coffee Board had wished to set a fLxed price for the crop year 1989. There are five expiration months on coffee futures contracts (March, May, July, September and December) and in December 1988 each had a price for delivery in these months in 1989. Based on historical cxport trcnds the Board could have predicted fairly well the proportion of the total exports available before each of these delivery months (e.g., January, February and March 10%; April and May 20%; June and July, 30%; August and September 25%; October, Novcmber and Deccmbcr 15%). These proportions could have then been used to obtain a weighted price for the coming year. This strategy was simulated using monthly data from 1980 to 1989 and the cffect on intra-year variability 27 observed by comparing the CVs of various price series. The results are reported in Table 8.1. for the world price, the threshold price (ten-year moving average), the domestic price A (the threshold price adjusted for subsidies and levies calculated using the world price), the fixed price (the weighted futurcs price) and the domestic price B (the threshold price adjusted for subsidies and levies calculated using the fixed price, instead of the world price, and assuming that the price band followed the moving average). -abl 8.J Intra-Year Meie Yariability 1921149. (CVsA Year World Threshold Domestic Fixed Domestic Pnce Price Price A Price Price B 1980 16.8 2.9 8.9 0.0 1.1 1981 6.2 2.0 4.1 0.0 1.0 1982 4.2 2.0 1.6 0.0 0.0 1983 5.5 1A 4.7 0.0 0.0 1984 2.7 1.6 1.9 0.0 0.0 1985 11.2 IS 7.4 0.0 0.7 1986 16.7 1.2 9.7 0.0 0.6 1987 83 2.5 3.9 0.0 0.0 1988 2.9 OS 2.5 0.0 0.2 1989 30.4 IA 16.0 0.0 0.7 Mean 10S 1.7 6.1 0.0 0.4 The inter-year CV of the world price is substantial, averaging 10.5% over the ten-year period. The instability in 1989 is clearly captured by the CV statistic, which reached 30.4% for that year. The domestic price A (based on the world price) also fluctuated throughout the period. The results showed that under the existing scheme the greater the intra-year variability of the world price, the greater it was for the domestic price for produccrs (e.g., in 1980, 1985, 1986 and 1989), even with subsidies and levies. In contrast, by using the futures contract to lock-in an external price, the inter-year fluctuations in the domcstic price are removed almost altogethcr. On average, the CV of the adjusted domestic price is 0.4%, indicating that produccrs are almost entircly insulatcd from intra-ycar price instability. The impact of opcrating this hedging strategy on intcr-vear price variability is reported in Table 8.2. 28 Tabl2 AeracE PIcesfor Stabilized Coffee Pre In pNO Us Dolams. Year Wodd Threshold Domestic Fund Fixed Domestic Fund Price Price Pice nsize Pri price size 1980 3466 2444 2953 68 4106 3273 73 1981 2869 2640 2749 82 2754 2697 95 1982 3088 2840 2964 87 2733 2733 96 1983 2911 3009 2898 89 2788 2788 96 1984 3189 3173 3157 91 3033 3033 97 1985 3231 3355 3203 91 2951 2985 96 1986 4295 3541 3897 104 3939 3740 100 1987 2S05 3421 2792 101 2792 3278 105 1988 3013 3257 3019 90 2926 2930 10S 1989 2387 3212 2684 85 2691 2791 101 Mean 3095 3089 3031 89 3120 302S 90 Std. Dev. 504.1 334.9 328.7 9.4 481.6 308.6 9.6 CV 16.2 10.8 103 10.6 H5A 10.2 8.9 The results show that the introduction of the hedging program has very little impact of inter-year variability compared to the stabilization fund. On average, the domestic price A (based on the world price) is $3,031 per ton, compared to $3,025 per ton for domestic price B (based on the fixed price). The stability of these prices was similar with coefficients of variation of 10.8% and 10.2%, respectively. The impact of the hedging strategy on the fund size is also shown in Table 8. 2. On average the mean of the fund with the hedging strategy is larger than that without it, while the variability of the fund is lower. This reflects the lower CV of the fixed price compared to the world price. This analysis shows that with a simple hedging strategy using futures contracts the Coffee Board can rix the cxtcrnal price it faces for an entire year at a time. This stability can be translated into a very stable within-year domestic prices received by producers. The clfcct of implemcnting this strategy on intcr-year variability is vcry small, however. Futures with longer maturities or rolling over futures may be used to managc this inter-year price risks. 29 9. Cmdushb This paper diswsed some of the financial instruments available to PNG for managing its external risks and illustrated their potential benefits. Given the importance and benefits of risk management, PNG should consider undertaking a risk management program. This could involve establishing a national financing and risk management strategy (authorized perhaps by PNG's National Borrowing Advisory Committce). Furthermore, the MRDC needs to be strengthened by establishing rules regarding not only the size of its participation in ncw projects, but also the type of external rnancing it uses (e.g., more commodity indexing features) and its risk management strategy. In the past, the MRSF has been well managed given the financial tools available. Howcvcr, the MRSF should now take advantage of the recent innovations in international capital markets by using instrunlenis stic h as commodity swaps to hedge its commodity price-sensitive tax revenues. This would allow stablc cash-llows to be passed on to the general budget, independent of commodity price movements. Other comnmioodity risk management techniques can also be introduced in managing the MRSP. With the expected expansion of the mineral and energy sector, PNG should look for ways to rcduce the variability of its economy through fmancial instruments, and, given its high international credit standing, is well placed to take advantage of longer-term commodity-linked instruments. 30 Coopers and Lybrand, (1989), Mining and Petroleum Taxation: A Guide for Operators and Cotractrs. Brogan, B. and J. Rewenyi, (Ed.), (1987), Commodity Price Stabili7ation in PNG - A Work-in-Progress Seminar. Australian Centre for International Agricultural Research, January. Coleman, J.R. and Y. Qian, (1991), 'Managing Financial Risks in PNG: Optimal Debt Portfolio," In review for PRE Working Paper Series, World Bank. Gemmill, G., (1985), 'Optimal Hedging on Futures Markets for Commodity-Exporting Nations", Europen Economic Review 27, pp. 243-61. Gilbert, C.L., (1990), "Domestic Price Stabilization Schemes for LDCs," mimeo, IECIT, World Bank. Guest, J., (1987), "Problems in Managing the MRSF", Bank of PNG Quarteriv Economic Bulletin June 1987. International Monetary Fund, (1990), 'Request for Stand-By Arrangement and Compcnsatory Financing of Export Fluctuations under the Compensatory and Contingency Financing Facility," March. Kolb, R.W., (1985), -Understanding Futures Markets, Scott, Foresman and Co. (London). Masuoka, T., (1990), "Asset and Liability Management in the Developing Countries. Modem Fmancial Techniques. A Primer," PRE Working Paper No. 454, World Bank. World Bank, (1989), Panua New Guinea OpDortunities and Challenges for Accelerated Development, Report No. 7707-PNG, April. World Bank, (1990), Papua New Guinea Strategy for Adjustment and Growth, Report No. 8518-PNG, April. 31 Debt Rntigp szm ~rggno C0Mggsitioin 0f PNG 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 PRlWCZPL RATI03 Total £xternal Debt (I) EDT/XGS 66.1 121.9 175.9 195.9 197.3 203.0 168.2 158.5 130.5 161.5 EDT/GNP 29.2 49.1 72.3 77.0 81.7 90.5 77.6 77.0 66.5 74.2 TDS/ MS 13.8 19.0 25.1 30.1 40.1 32.7 26.8 24.6 26.6 34.3 IHTMX0S 8.6 10.0 12.7 13.8 15.8 13.5 11.3 11.1 10.0 11.2 INT/GNP 2.9 4.0 5.7 5.4 6.6 6.0 5.3 5.4 5.1 5.1 Cancessional/EDT 12.2 9.7 8.6 12.0 14.0 16.3 20.4 24.1 19.3 19.1 1980 1981 1982 1963 1984 1985 1988 1987 1988 1989 CURMBWC'Y CCFEOSITrON OF DEBT OUTSTANDINO (PFfCERI) U.8.Dollcrs 35.0 27.9 39.0 48.4 45.5 39.0 35.7 29.2 28.9 31.9 Mixed Currency 27.4 41.3 36.2 31.1 29.0 28.9 28.0 27.6 29.9 31.0 Japanese Yen 4.4 3.7 3.4 3.0 8.4 14.2 16.1 18.0 20.8 23.0 Deutsche Mark 5.2 3.9 3.1 1.9 2.0 2.1 2.3 2.3 1.8 1.8 French Franc .0 .0 .0 .0 .2 .4 ,3 .3 .3 .2 Pound Sterling 2.7 2.3 2.4 1.7 1.2 1.3 1.9 2.1 2.0 1.5 Swiss Franc 25. 4.5 3.4 2.5 4.2 7.2 9.5 10.3 9.6 6.1 Other Currencies 19.6 16.4 12.4 13.4 11.7 9.4 9.0 13.3 10.2 9.6 Exchoane Rate Val Effects /a 0 -12 -20 -13 -20 40 87 140 -44 -50 Notes: a/ IDP. debt valuation effects are shown from 1985 onward, millions of dollars. PRE Working Paper Series Contact 312 AlAhar Date fAuthr bLaoer WPS730 Wage and Employment Policies in Luis A. Riveros July 1991 V. Charles Czechoslovakia 33651 WPS731 Efficiency Wage Theory, Labor Luis A. Riveros July 1991 V. Charles Markets, and Adjustment Lawrence Bouton 33651 WPS732 Stabilization Programs in Eastern Fabrizio Coricelli July 1991 R. Marlin Europe: A Comparative Analysis of Roberto de Rezende Rocha 39065 the Polish and Yugoslav Programs of 1990 WPS733 The Consulting Profession in Syed S. Kirmani July 1991 INUDR Developing Countries: A Strategy for Warren C. Baum 33758 Development WPS734 Curricular Content, Educational Aaron Benavot July 1991 C. Cristobal Expansion, and Economic Growth 33640 WPS735 Traditional Medicine in Sub-Saharan Jocelyn DeJong July 1991 0. Nadota Africa: Its Importance and Potential 31091 Policy Options WPS736 Wages and Employment in the Simon Commander August 1991 0. Del Cid Transition to a Market Economy Fabrizio Coricelli 39050 Karsten Staehr WPS737 External Shocks, Adjustment Delfin S. Go August 1991 A. Bhalla Policies, and Investment: Illustrations 37699 from a Forward-looking CGE Model of the Philippines WPS738 Tax Competition and Tax Ravi Kanbur August 1991 J. Sweeney Coordination: When Countries Michael Keen 31021 Differ in Size WPS739 Managing Financial Risks in Papua Jonathan R. Co;eman August 1991 J. Carroll New Guinea: An Optimal External Ying Oian 33715 Debt Portfolio WPS740 The Onchocerciasis Control Program Beinhard H. Liese August 1991 0. Nadora in West Africa: A Long-term John Wilson 31091 Commitment to Success Bruce Benton Douglas Marr WPS741 When Does Rent-Seeking Augment David Tarr August 1991 D. Ballantyne the Benefits of Price and Trade Reform 37947 on Rationed Commodities9 Estimates for Automobiles and Color Televisions in Poland PRE Working Paoer Series Contact Th1Q 1~Author D for pape WPS742 The Cost of the District Hospital: A. J. Mills August 1991 0. Nadora A Case Study from Malawi 31091 WPS743 Antidumping Enforcement in the Angelika Eymann August 1991 N. Artis European Community Ludger Schuknecht 37947 WPS744 Stainless Steel in Sweden: Anti- Gunnar Fors August 1991 N. Artis dumping Attacks Good International 37947 Citizenship WPS745 The Meaning of 'Unfair in U.S. J. Michael Finger August 1991 N. Artis Import Policy 37947 WPS746 The Impact of Regulation on Dimitri Vittas August 1991 Financial Intermediation WPS747 Credit Policies in Japan and Korea: Dimitri Vittas August 1991 A Review of the Literature WPS748 European Trade Patterns After the Oleh Havrylyshyn August 1991 N. Castillo Transition Lant Pritchett 37947 WPS749 Hedging Commodity Price Risks in Stijn Claessens August 1991 S. Lipscomb Papua New Guinea Jonathan Coleman 33718 WPS750 Reforming and Privatizing Poland's Esra Bennathan August 1991 B. Gregory Road Freight Industry Jeffrey Gutman 33744 Louis Thompson