LONG-TERM OIL WARRANTS: AN APPLICATION TO VENEZUELAN DEBT RELIEF by THOMAS CEPHIS WILLIAMS A.B., Harvard College (1982) Submitted to the Alfred P. Sloan School of Management in Partial Fulfillment of the Requirements for the Degree of MASTER OF SCIENCE in Management at the MASSACHUSETTS INSTITUTE OF TECHNOLOGY June 1990 o Massachusetts Institute of Technology (1990) ALL RIGHTS RESERVED Signature of Authori u MIT Sloan Schiooi f Miigmet May 18, 1990 Certified by Henry D. Jacbby Professor of Management Thesis Supervisor Accepted by ! -w Jeffrey A. Barks Associate Dean, Master's and Bachelor's Program 1 MASSACHUSETITNSS TITUTE OFT ECHNOLOGY JUL 06 1990 LIBRARIES ARCHIVES LONG-TERM1 OIL WARRANTS: AN APPLICATION TO VENEZUELAN DEBT RELIEF by Thomas C. Williams Submitted to the Alfred P. Sloan School of Management on May 18, 1990, in partial fulfillment of the requirements for the Degree of Master of Science in Management ABSTRACT Long-term warrants have been proposed as part of packages for restructuring the debt of highly-indebted countries (HICs), such as Venezuela. The proposed warrants are essentially call options on a commodity produced by the countryj (e.g., oil). Valuation of these warrants is problematic, since widely-used methods for derivative- asset valuation, such as the Black-Scholes formula, become less precise as time to maturity increases. A ten-year strip of one-year warrants is used, with Venezuelan crude oil as the underlying asset. Results from a standard Black-Scholesv aluation for the warrant strip are compared with results from a Jacoby-Laughton/Fama-Cox-Ross derivative-asset valuation model using the same set of asstnptions about the economic environment (inflation rate, real interest rates, etc.) and asset-related factors (e.g., initial oil price, volatility of oil prices). The theoretical foundations of the Jacoby-Laughton model come from the work of Fama (which led to the Capital Asset Pricing Model, or CAPM), and the Cox-Ross options-pricing model. The Jacoby-Laughton/Fama-Cox-Ross (JL/FCR) model produced the same results as Black-Scholes when given the same parameters. The JL/FCR model permits correction for the Black-Scholest endency to overvalue long- term options, as well. A method of measuring the political risk of the warrants was developed, and the amount of debt relief gained from a warrant-based restructuring was calculated. More accurate valuation of such long-term options may lead to greater flexibility in restructuring HIC long-term debt. Thesis Supervisor: Prof. Henry D. Jacoby Title: Professor of Management 2 Acknowledgements I would like to thank Professors Henry D. Jacoby and Donald R. Lessard of the MIT Sloan School, and Professor David G. Laughton of the University of Alberta, Canada, for their invaluable aid over the past semester. This work most definitely would not have been possible without Professor Jacoby's firm guidance, Professor Lessard's fertile mind and Professor Laughton's deep knowledge of this subject. I also want to thank my fiancee, Cheryl E. Houston, for her unflagging support during the sometimes trying process of this work's development. Without Cheryl, I would not have even begun. 3 Table of Contents 1 Introduction ................................... 5 2 Valuation Issues in LDC Debt Reduction .... ......... ........ ...·... 7 2.1 History of Debt-Reduction Efforts. ..·... .. ·.. .. .... ... ... 7 2.1.1 Mexico............... ........................ ....8....... 2.1.1.1 The 1989 Agreement (exit bond) ............... 9 .......... ............... 10 2.1.2 Venezuela............. 2.1.2.1 The 1989 Bond/Warr nt Proposal ............. 11 2.1.2.2 The 1990 Venezuelan Agreement .............. 12 2.2 Valuation Problems ............ ....................... .13 2.3 Political Risk ................ ....................... .14 3 Valuation Methods ............................................ 20 3.1 Notation .......................................... 20 3.2 Black-Scholes Method ................................. 21 3.3 Fama-Cox-Ross Method ............................... 23 4 Setup for the Valuation Experiment ............. .................. .29 .................. .30 4.1 Environmental Assumptions .......... 4.2 Oil Price Assumptions ............... .................. .30 4.3 Valuation of the Underlying Asset ...... ................ 31... 4.4 Transformations for Shorter Time Periods .................. 31 .................. .33 5 Venezuelan Example ........................ .................. .34 5.1 Black-Scholes Analysis ............. .................. .37 5.2 Basic Fama-Cox-Ross Analysis ........ ................. . 37 5.2.1 FCR Model Validation ......... .................. 40 5.2.2 Debt-Capacity Analysis ......... .................. ..41 5.2.3 Effects of Political Risk ......... 5.3 Alternative Specification of the InformatioiinM odel ........... 44 6 Sumnmary 50 . . . . ·········· .········· . . . .. . . . . . . . . . .. 6.1 Possible Applications ....................... 50 6.2 Summary of Results ....................... 50 4 1 Introduction Large sums of money were borrowed from commercial banks by countries in the developing world in the 1970st o fund natural-resource development projects, such as oil production. These projects were often conceived during times of high market prices for the commodity, and expected returns on the borrowed money clearly exceeded the cost of borrowing (Randall 1987; Teichman 1988). In the 1980s,o il prices fell to around $14 U.S. per barrel, resulting in severe cash-flowp roblems for oil-producing countries such as Mexico and Venezuela. Oil production was the main revenue source for these countries, yet oil revenue was insufficient to meet the countries' debt payments. Both Mexico and Venezuela have sought debt relief from their lenders through renegotiation of existing debt. The lenders seek to maximize the amount repaid while minimizing their exposure to risk of non-payment. The countries want to minimize the imnpacot f debt service while retaining access to loans. Lenders and countries both want to continue a valuable relationship, as recent debt-restructuring agreements in both Mexico and Venezuela clearly show. Current debt-relief efforts have been heavily influenced by the Brady plan, a framework for HIC debt restructuring proposed and supported by the United States. The Brady plan (McNamee et al. 1990) emphasizes negotiations between the debtor governments and their creditor banks to reduce the amount of existing debt and/or the size of debt payments, with potential supplementary lending from supranational organizations such 5 as the World Bank and the IMF. Brady plan-style agreements do not match the amount of the government's debt payments to the ability of the government to pay as a function of time. This is a significant flaw in this type of solution to debt relief. This thesis will concentrate on the financial aspects of proposed solutions to the Venezuelan debt-reduction problem, with special focus on a 1989 proposal based on a bond/warrant combination. A valuation method, based on the Jacoby-Laughton derivative asset valuation method, will be developed for the warrants described in the 1989 proposal. Finally, a measure of the warrants' political risk will be derived from the valuation method. 6 2 Valuation Issues in LDC Debt Reduction Determination of the real value of debt, of debt reduction and of risk is one of the major problems seen in debt-reduction negotiations. Debtors and lenders have opposing goals in valuation; debtors seek to gain the mximum credit at the lowest cost, while lenders seek to maximize return on funds lent while minimizing risk. This section will review the recent history of debt-reduction efforts in Mexico and Venezuela as a framework for understanding some of the political and financial issues in valuation of LDC debt relief. 2.1 History of Debt-Reduction Efforts During the late 1970s and early 1980s,d eveloping countries saw decreases in GNP and government revenue, as world prices for several important commodities declined. Oil was probably the most important of these commodities for fast-growing LDCs, which were either oil exporters (Mexico, Venezuela) or oil importers (Brazil). Heavy government borrowing for projects in the 1960s and 1970s was based on the expectation of higher commodity prices in the 1980s than actually occurred. This brought many'debt crises' to LDC governments, as they were less able (or unable) to meet their debt payments. As the risk of default increased, negotiations were initiated to restructure foreign debts in countries ranging from Mexico to the Philippines. The current approaches to the debt-relief problem in Mexico and Venezuela are similar, and the following sections will describe the historical context behind these countries' 7 methods in greater detail, with emphasis on recent debt-reduction negotiations. 2.1.1 Mexico The deterioration of Mexico's economy during the 1960s and early 1970s decreased government revenue and provoked increasing dissatisfaction with the Mexican government's economic policy. The 'shared development' strategy, which shifted government investment away from the industrial sector and towards agriculture and social-welfare programs, was a political solution to the government's dilemma. Mexico's substantial petroleum reserves were increasingly seen as the source of the financial solution to the government's problem, but exploitation of these reserves for export ran counter to deeply-held beliefs, dating to the Mexican revolution, about the proper use of Mexican natural resources. Higher levels of oil production required an increase in investment in oil-production capacity. Mexican production capability in the early 1970s was insufficient to meet Mexican demand, so substantial investment was needed to support an export-oriented strategy. This change in oil-development strategy laid the foundations for the 1982 'debt crisis'. Increased dependence on oil exports required the government to continually increase its capital investments for exploration and production. Increased need for capital investment further accelerated Mexico's acquisition of foreign debt, as Mexico was not 8 able to finance both increased oil production and increased social spending out of its current income. The combination of falling prices for crude oil in the early 1980s and increasing short- term world interest rates led to the 1982 debt crisis. As crude prices dropped, Mexico was increasingly forced to seek short-term bank loans to make payments on its huge foreign debt (about $84 billion in 1982). Foreign banks became less willing to make short-term loans to Mexico as its oil revenues declined, and loan terms were made more restrictive (shorter loan periods, higher interest rates). 2.1.1.1 The 1989 Agreement (exit bond) The 1982 debt crisis was resolved through a debt-restructuring agreement which allowed lenders to wind down their lending to Mexico through acceptance of one of two types of "exit bond", or by refinancing the bank's loan through a combination of new debt, or "new money', and existing debt ("old money'). '"Newd ebt" loans were limited to an additional amount equal to 25 percent of the existing debt owed to a particular bank. When existing debt was swapped for "exit bonds", Mexico's liability to the lenders was reduced by either refinancing the same amount of debt at a lower interest rate ("interest- reduction bond"), or by reducing the principal amount of the debt ("principal-reduction bond") while keeping the old, higher interest rate. The "exit bond" option allowed lenders pessimistic about Mexico's future prospects to avoid a total loss from their Mexican debt exposure, while providing some relief to the Mexican government by 9 reducing its payments to the lenders. Refinancing with "new money' was chosen by lenders more optimistic about Mexico's long-term ability to repay both principal and interest on its debt. The agreement allowed the banks to structure their Mexican debt portfolio, using these three basic building blocks in whatever proportion made sense to the bank. No consistent assessment appears to exist of the riskiness of the restructured debt; some lenders are writing down the principal-reduction bonds and the interest- reduction bonds, (e.g., J.P. Morgan), while others (e.g. Manufacturers Hanover) are writing down only the principal-reduction bonds (American Banker 1990). The flexibility of the Mexican restructuring plan makes Mexican-style restructurings attractive to both lenders and governments, and the plan has served as a model for refinancing proposals in other countries. The followings ection will discuss the historical context of the recent Venezuelan debt-relief negotiations and two recent restructuring proposals (one based on a bond/warrant combination and the other a Mexican-style proposal). 2.1.2 Venezuela Revenue from oil production is very important to Venezuela. Oil export revenue is a very large percentage (over 90 percent through the 1970s) of all Venezuelan revenue received from exports. This great dependence on one source of income from a non- renewable resource has been of great concern to recent Venezuelan governments, which have sought to direct this income towards investments in non-petroleum renewable resources and general infrastructure improvements. Investments in the petroleum sector 10