UNITED GRAIN GROWERS: ENTERPRISE RISK MANAGEMENT AND WEATHER RISK Scott Harrington Greg Niehaus ABSTRACT In August of 1999, Mike McAndless, the risk manager of United Grain Growers (UGG), was preparing for a meeting with the firm's chief financial officer, Peter Cox. Mike and Peter had spent considerable time over the past three years with representatives of the Willis Group Ltd., a large international insurance broker, identifying and measuring UGG's major sources of risk. The risk assessment process indicated that, although UGG hedged most of its currency and com- modity price risk and purchased insurance against property and liability losses, the firm's earnings still exhibited substantial volatility. This volatility was, in large part, due to the weather. Mike and Peter had to decide whether to re- tain the risk or shift it to another party using one of two innovative contractual arrangements: weather derivatives or a new type of insurance contract. INTRODUCTION This is a case study designed to be used in a risk management class to illustrate enterprise risk management. This case does not reveal United Grain Grower's (UGG) decisions. Instead, the case ends with seven questions for students to answer. COMPANY BACKGROUND Based in Winnipeg, Manitoba, UGG provides commercial services to farmers and mar- kets agricultural products worldwide. It was founded in 1906 as a farmer-owned coop- erative, and became a publicly traded company on the Toronto and Winnipeg stock ex- changes in 1993. Figure 1 provides information on UGG's stock price since going public. Although UGG is a public company, it retains some of its farmer cooperative roots. The company has both members and shareholders. At the time of the initial public offering, the members of the cooperative (farmers) automatically became members of the new organization, and they also received limited voting common shares (thus making them both members and shareholders of the new organization). An individual, who is not currently a member, can apply for membership if the individual does a minimum amount of business with the company. The initial public offering, as well as subsequent equity offerings, allowed nonmembers to become shareholders. Although a member is not entitled to share in any profit or distribution by the company (unless the member is also a shareholder), members have control rights. Of the 15 peo- ple on UGG's board of directors, 12 must be "members" who are elected by delegates representing members from various geographical regions. Business Segments UGG is comprised of four main business segments: Grain Handling Services, Crop Production Services, Livestock Services, and Business Communications. As illustrated in Figure 2 and discussed below, UGG's four business units help farmers plan, produce, and market their products. Western Canada is a major producer and exporter of wheat, barley, canola, and other grains and oilseeds. The role of UGG's Grain Handling Services unit (comprised of Farm Sales and Services, Marketing and Transportation Services, and Terminal Services divisions) is to identify sources of grain and oilseeds and deliver them to exporters and to domestic end users, such as food processors. A farmer's production of grain and oilseeds usually is transported to a country elevator, where the product is weighed, graded, blended, purchased, and stored. From the elevator, the product is shipped to a domestic consumer (e.g., a mill) or to an export terminal. UGG historically owned hundreds of relatively small "country" elevators, which the firm has been replacing with a smaller number of large, high throughput, more efficient Note: Diagram illustrates UGG's four business segments: Crop Production Services, Livestock Services, Grain Handling and Marketing, and Communications/Information and the planning, production, and marketing services that each of the segments provide to farmers. Documents provided to the authors by UGG. elevators. The map of western Canada in Figure 3 identifies the locations of UGG's main elevators and export terminals. The farming industry in Canada is regulated by several government agencies. The Canadian Wheat Board (CWB) markets human consumable grains on behalf of farmers. About 85 percent of the wheat and 45 percent of the barley produced in Canada is sold through the CWB. The CWB must ensure that the sales it has arranged are available to customers at the agreed upon site and date. Thus, the CWB contracts with companies like UGG to collect, store, and deliver grains. About 60 percent of UGG's grain handling unit's business is on behalf of the CWB. The prices paid to farmers and the prices for storage and transportation of "board grains" are determined by the CWB. The Canadian Grain Commission regulates grain handling and maintains quality stan- dards for Canadian grain. Firms like UGG must obtain an operating license from the Commission. The Commission also maintains extensive records of the grain that is shipped from country elevators and from export terminals. Table 1 provides data on grain shipments and deliveries for the industry and for UGG from 1981 through 1999. UGG's competitors in the grain handling business are listed in Table 2 along with approx- imate market shares in 1999. UGG's market share of approximately 15 percent makes it the third largest provider of grain handling services in western Canada. Table 3 provides information on the volume of grain shipped by UGG, as well as UGG's gross margin and earnings on grain shipments. The table also provides information on gross margin and earnings per tonne of grain shipments.

Note: White area indicates the area served by UGG, dots indicate the location of high throughput grain elevators, and the squares indicate the location of export terminals. Documents provided to the authors by UGG. The Crop Production Services unit provides inputs (e.g., seed, fertilizer, and crop pro- tection products) to farmers. In addition, through its Farm Sales and Services division, it provides a range of consulting, agronomic, and financial services to farmers. UGG tries to differentiate itself from its many competitors by developing distinctive products sold under brand names and by the provision of superior services to farmers. UGG's third largest unit is Livestock Services, which provides inputs to producers of cattle, hogs, and poultry. This unit also faces competition from a number of other grain and feed companies. UGG's smallest business unit is Farm Business Communications, which provides information needed to run a profitable agribusiness. In addition to pub- lishing periodicals (Farm Investor Newsletter and Disease, Weeds & Insects), this unit has developed Web-based information on weather, market prices, and agribusiness news. Figure 4 illustrates earnings before interest and taxes (EBIT) for each of UGG's business units over time. The two largest lines of business, Grain Handling Services and Crop Production Services, account for over 80 percent of UGG's earnings in most years. The figure also illustrates the substantial earnings volatility in these main business segments.

Financial Results Grain Handling Crop Production Services Livestock Services Bu siness Communications Table 4 contains information from UGG's balance sheet, income, and cash flow state- ments. Earnings before interest, taxes, depreciation, and amortization (EBITDA) declined substantially in 1999 relative to the prior years. UGG increased capital expenditures sub- stantially in 1998 and then again in 1999. Most of these expenditures were for large high throughput grain elevators. As a result of the low EBITDA in 1999, UGG's return on equity (defined as net earnings to book value of equity) was just 1.17 percent. Note as well that in 1999, the percentage of the firm's total assets financed with debt increased to about 37 percent with the issuance of another 50 million Canadian dollars in long-term debt.

CORPORATE RISK MANAGEMENT Background Restated 1994 $156,030 25,538 12,612 3,772 153 12,533 27,725 $75,028 153,228 564,043 140,516 59.11% 0.06% $0.01 1.30 1995 $185,637 30,573 15,151 282 $7,385 16,177 43,894 $44,573 182,079 544,284 130,620 57.72% 2.20% $0.24 1.47 1996 $198,749 40,198 24,090 8,065 5,851 21,322 26,826 $71,557 190,308 531,416 133,694 55.36% 4.30% $0.45 1.94 1997 $216,260 54,788 38,452 24,744 9,059 32,770 21,904 $101,790 193,323 489,214 161,290 36.01% 8.51% $0.89 2.66 1998 $224,953 60,577 43,335 31,926 16,332 35,871 53,760 $136,155 226,304 515,209 234,611 26.24% 8.69% $0.91 2.08 1999 $209,227 42,423 21,636 8,067 3,575 29,853 91,002 $119,249 287,442 554,322 233,182 36.76% 1.17% $0.15 1.72 For many corporate risk managers, "risk management" refers to the management of so- called "pure risks," e.g., losses from property damage, liability suits, and worker injuries. These risks typically are managed through a combination of loss control (efforts to reduce the likelihood or magnitude of losses) and loss financing through internal retentions or the purchase of insurance.

200 RISK MANAGEMENT AND INSURANCE REVIEW In the 1980s and 1990s, a different type of risk managementfinancial risk manage- mentgrew in importance at many corporations. Financial risk management typically refers to the management of price risks, e.g., losses from changes in prices, such as exchange rates, interest rates, commodity prices, and credit exposures. These risks usu- ally are managed through derivatives contracts, such as options, forwards, futures, and swaps. In most corporations, financial risks were managed separately from pure risks, and the terminology and methods used by managers of financial risk differed from those used by managers of pure risk.

Enterprise Risk Management During the latter part of the 1990s, some managers started to question the desirability of managing pure risk and financial risk separately. They also began to consider risk exposures that were not handled by pure risk or financial risk managers. For example, a firm might have operational risks that were being ignored by the risk managers because there was not an established contract (insurance or derivative) that could be used to shift the risk to another party. The idea that a firm should examine all of its risk exposures and deal with them using a consistent framework came to be known as enterprise risk management (ERM). To facilitate communication among different areas within a firm and the adoption of a consistent risk management framework, some firms even established a new positionthe chief risk officer.

ENTERPRISE RISK MANAGEMENT AT UGG Several factors led UGG to investigate enterprise risk management. One factor was that the Toronto Stock Exchange directs the board of directors of all listed corporations to identify the corporation's principal risks and to implement appropriate systems to manage these risks. Other factors included increased requirements for disclosure of risk exposures, increased emphasis on risk management by credit rating agencies, and UGG's perception that equity analysts recommendations were sensitive to earnings results that deviated from forecasts.

Identifying and Quantifying Risk Exposures UGG started by forming a risk management committee, consisting of the CEO, CFO, risk manager, treasurer, compliance manager (for commodity trading), and manager of corporate audit services. This committee, along with a number of UGG employees, then met with a representative from Willis for a brainstorming session to identify the firm's major risks. This process identified 47 exposure areas, from which six were chosen for further investigation and quantification. The six risks were (1) environmental liability, (2) the effect of weather on grain volume, (3) counterparty risk (suppliers or customers not fulfilling contracts), (4) credit risk, (5) commodity price and basis risk, and (6) inventory risk (damage to products in inventory). Willis Risk Solutions, a unit of the Willis Group Ltd., took on the task of gathering data and estimating the probability distribution of losses from each of the six risk exposures. These probability distributions were then used to quantify the impact of each source of risk on several measures of UGG's performance, including return on equity, economic value added, and earnings before interest and taxes (EBIT). Figure 5 provides an example of the type of analysis conducted by Willis Risk Solutions. The example is based on UGG's counterparty risk. Based on data provided by UGG and discussions with UGG employees, Willis estimated that the number of counterparty losses per year could be described by a Poisson distribution (see Part A in Figure 5) and that the loss severity on any given loss could be described by a lognormal distribution (see Part B). Given the probability distributions for the number of losses and for the loss per event, an annual loss distribution from counterparty risk could be estimated (see Part C). Finally, the impact of counterparty risk on the probability distributions of various performance measures (e.g., EBIT) could be estimated under the assumption that all other risk factors took on a specific value (see Part D).1 The analysis conducted by Willis Risk Solutions led to the conclusion that, of the six risks originally identified, UGG's main source of unmanaged risk was from the weather. The parties therefore focused their energies on understanding how weather affected UGG's performance. Ken Risko and Michelle Bradley, a statistician and an actuary, re- spectively, for Willis Risk Solutions, conducted an in-depth regression analysis of how crop yields in each province of western Canada were influenced by temperature and precipitation. Examples of the regression analysis conducted by Ken and Michelle are presented in Table 5. The table provides the results of estimating a regression equation where the dependent variable is the crop yield (bushels per acre) for either wheat or oats, and the explanatory variables are a time trend (to capture productivity increases over time), the average June temperature, and the average July precipitation. A large number of different combinations of weather variables were tried and these two variables provided a good parsimonious model for explaining crop yields. The analysis was conducted using data from 1960 to 1992 for the provinces of Alberta, Manitoba, and Saskatchewan. Similar analysis was also conducted for other grains and seeds. To illustrate the results, consider the first row of Table 5. The positive and statistically significant coefficient on the time trend variable indicates that Alberta wheat yields have increased over time. The negative and statistically significant coefficient on the average June temperature variable indicates that wheat yields in Alberta are negatively related to the average June temperature. Finally, the positive coefficient on the average July precipitation variable indicates that crop yields increase on average with rainfall in July. The R-squared indicates that about 68 percent of the annual variation in Alberta wheat yields is explained by these three variables. The remainder of Table 5 indicates that, in general, crop yields for wheat and oats have increased over time, are negatively related to average June temperature and positively related to average July precipitation. There are, however, some exceptions to these gen- eralizations. The exhibit also indicates that the three variables in the regression equation explain a substantial proportion of the variability in yields in all of the provinces, i.e., the R-squareds generally are high. The regression results can be used to assess how expected crop yields would be af- fected by deviations from normal weather conditions. For example, if temperature and precipitation were expected to take on their historical average values (presented in Table 6), then the predicted Alberta wheat crop yield for 2000 would be Yield = 59.88 + .33(40) 0.76(56.6) + 2.7(2.06) = 35.6 bushels per acre. If instead the average June temperature was higher than the mean value by one standard deviation (2.2 degrees from Table 6), the Alberta wheat crop yield would be predicted to be Yield = 59.88 + .33(40) 0.76(58.8) + 2.7(2.06) = 34.0 bushels per acre. Having established a relationship between crop yields and weather, Ken and Michelle then estimated the relationship between crop yields and UGG's grain volume. They first calculated a weighted average crop yield for western Canada using crop yields by grain/seed and by province and the proportions of total production of each grain/seed in each province. The values for this weighted average crop yield are reported in Table 1. They found that UGG's grain volume in year t was highly correlated with overall crop yields in year t 1.

The next step in Ken and Michelle's analysis was to relate UGG's grain volume to UGG's financial results using the information in Table 3. For each tonne of shipments, UGG had gross profit of 21.2 Canadian dollars on average from the 1997 through 1999. To summarize, Ken and Michelle established a relationship between weather and UGG's gross profit using the following steps and information: Weather Crop Yields UGG's Grain Volume UGG's Profit Table 5 Table 1 Table 3 They illustrated their results by graphing UGG's actual gross profit and what gross profit would have been if the effects of weather were removed. Their graph is reproduced as Figure 6. ALTERNATIVE RISK MANAGEMENT APPROACHES Having quantified their exposure to weather risk, UGG had to decide what to do about it. They explored several options. Retention One approach was to continue operating as they had been and not try to reduce their weather exposure. As previously discussed, this approach exposed their profitability to large swings due to weather variation. There were several disadvantages of such volatility. First, UGG had been and planned to continue making large investments in storage fa- cilities (grain elevators). The ability to finance these capital expenditures from internally generated funds would allow the firm to avoid the costs associated with raising exter- nal capital. And, to the extent that external capital would be needed, the rate that the firm would have to pay on borrowed funds would likely be higher if they retained the weather risk. Second, the variability in its cash flows caused UGG to use equity capital as a cushion against unexpected low cash flows in any given year. If the firm could reduce its weather risk, it could increase the proportion of the firm financed with debt without paying higher yields, which in turn would allow it to gain additional interest tax shields. Third, although much of UGG's current business could be characterized as a commodity business, UGG tried to distinguish itself from competitors by creating products with brand names and by providing on-going services to customers. Stability in the firm's cash flows would help the firm characterize itself as a company that suppliers and customers could rely on for service and high quality products for many years. Moreover, the importance of supplier and customer relationships was likely to increase in the coming years as the marketplace for agricultural products adjusted to scientific advances. Analysts predicted that over the next decade, food producers would demand specific genetically engineered crops, which in turn would require farmers to plant specific seeds. The coordination of these activities between farmers and food producers would require an information, storage, and transportation network. UGG saw itself as a provider of these intermediary services. The main advantage of retaining the weather risk was the cost associated with shifting it to someone else. In addition, Mike and Peter were not sure that the capital markets really would reward the firm for eliminating weather risk, given that this was a risk that most investors could easily diversify on their own.

Weather Derivatives Unhedged Profits 80 80 60 80 60 12345678 9 10 11 Weather Index Payoff on a Weather Derivative 1 2 3 4 5 6 7 8 9 10 11 Weather Index Hedged Profits 1 2 3 4 5 6 7 8 9 10 11 Weather Index In the late 1990s, weather derivatives were a relatively new risk management tool. These contracts were sold in the over-the-counter (OTC) market by firms such as Enron. A contract could be tailored on a number of dimensions to meet the specific needs of the buyer. For example, the underlying variable determining the payoffs could be one or a combination of weather variables, such as average temperature, rainfall, snowfall, a heat index, or the number of heating or cooling degree days. The payoff structure could resemble a put option, a call option, a swap, or combinations of these structures. Figure 7 provides an example of how UGG could potentially use a weather derivative. Suppose that, based on Willis's analysis of the sensitivity of crop yields to weather and the sensitivity of gross profit to crop yields, UGG's expected gross profit exhibited a pattern depicted in Part A of Figure 7. The vertical axis measures expected gross profit and the horizontal axis measures a weather index, which equals a weighted average of various temperature and precipitation measures in western Canada. As the index increases, expected gross profit increases (because crops' yields increase, which in turn increases UGG's shipments of grains and seeds). For simplicity, the illustration assumes that the relationship between gross profit and the weather index is linear. Since low values of the weather index correspond to low expected profits for UGG, a derivative contract that would pay UGG money when the index is low would provide a hedge. For example, the put option structure illustrated in Part B in Figure 7 would help to hedge UGG's risk. When the put option payoff from Part B is added to expected gross profit from Part A, UGG's expected gross profit would vary with the weather index as depicted in Part C. Hedging their weather risk with derivatives was feasible, but it suffered from several dif- ficulties. Although Willis had performed a sophisticated analysis of the effect of weather on UGG's gross profit, the results of this analysis had to be converted into a desired contract structure. That is, the underlying weather index that determined the derivative contract's payoff would need to be specified. Next, the effectiveness of the derivative contract in hedging UGG's risk would have to be assessed. UGG then would have to obtain price quotes in a marketplace that had relatively few participants. The Insurance Contract Idea When discussing the weather analysis, Mike McAndless and Peter Cox thought of an alternative way of dealing with the firm's weather risk. They knew that the primary reason weather was important was because weather affected UGG's grain shipments. They therefore wondered whether they could construct an insurance contract that would pay UGG when its grain shipments were abnormally low. The obvious problem with such a contract is moral hazardUGG's pricing and service also influences its grain shipments. One solution to this problem was to use industry-wide grain shipments as the variable that would trigger payments to UGG. Industry shipments would likely be highly correlated with UGG's shipments, which would imply that the basis risk would be minimal. In addition, because of its relatively low market share, UGG would have minimal effect on the value of industry-wide shipments, which would significantly reduce the moral hazard problem. Mike and Peter also considered the possibility of integrating grain volume coverage with UGG's other insurance coverage. Currently, UGG purchased a number of different insurance policies for various traditional risk exposures. For example, they purchased a variety of policies to cover their property exposures (e.g., a boiler and machinery policy to cover losses on machinery and equipment) and liability policies to cover their exposure to tort liability (e.g., environmental impairment liability). Each policy had its own retention level and its own coverage limit. By integrating its various coverages under one policy, UGG could replace the individual deductibles and limits with an overall annual aggregate deductible and limit that would apply to all or a subset of losses, including grain volume losses. Mike called Willis and asked them to investigate the possibility of structuring an insur- ance contract on industry grain shipments. Willis then contacted several major commer-cial insurers, including a division of the large reinsurer Swiss Re, called Swiss Re New Markets. Located in New York, this group structured innovative risk financing deals for commercial entities. In preparation for a meeting with a group from Swiss Re New Markets, Mike and Peter wanted to answer the following questions: (1) Given that any method of reducing the weather risk exposure will be costly, what are the benefits to UGG's diversified owners from reducing the weather risk? (Hint: What characteristics of UGG's operations and strategy would make risk reduction potentially beneficial?) (2) Should UGG's rather unique ownership structure influence the decision to reduce the weather risk exposure? (3) How could the parties structure a weather derivative to cover the exposure? More specifically, what would be the underlying index? Would they need a separate contract for each crop and each province? (Hints: In constructing an underlying index use the regression analysis in Table 5. In discussing whether they would need a separate contract for each crop and province, consider basis risk and also transaction costs.) (4) How could the parties structure an insurance contract to cover the grain volume exposure? More specifically, how would a loss be defined? And, what would be the payment to UGG conditional on a loss? (Hint: Use information in Tables 1, 2, and 3.) (5) Whataretheadvantagesanddisadvantagesofintegratingthegrainvolumecover- age with the firm's other insurance coverages? That is, instead of having separate policies with separate deductibles and limits for the various exposures (includ- ing the grain volume exposure), what are the advantages and disadvantages of bundling all of the firm's exposures in one policy with one deductible and one limit? (6) Ignoring cost differences, are there any advantages of the insurance contract ap- proach versus the use of weather derivatives? (Hint: Comment on basis risk and moral hazard.) (7) Are there any loss control measures that could be used to manage UGG's weather risk?

Questions to Answer

1. What are the sources of UGGs revenues and costs and how are they related to their risks?
2. Why should UGG worry about these risks if Owners can diversify away from these risks? Owners can just purchase a well-diversified portfolio to avoid these risks can they not?
3. Do you think an insurer would be more willing to write a contract with UGG that pays UGG if the firm's grain handling division profits are below a certain level or a contract that pays them money if Canadian grain volumes are below a certain level? Why?
4. How important is weather risk to the firm relative to other risks? How would you go about assessing the effect of weather risk on the firm? What affects grain volume?
5. The ideal contract here is a profit protection contract. Why won't an insurer sell UGG a contract that says, pay us a premium of x and we will pay you if the difference between your expected profit and what you actually receive in profit?