Financial risk monitoring and transaction costs in coffee & soybean trading companies and processors

Financial risk monitoring and transaction costs in coffee & soybean trading companies and processors

Edson Costa Bignotto^I; Adriano Azevedo Filho^II

^IM.Sc.Economics, ESALQ-USP; Financial Analyst — MAPS Risk Management Solutions; [e-mail]

^IIPhD in Economics; Assistant Professor, Department of Economics, Business Administration and Rural Sociology at ESALQ-USP. Piracicaba - SP - Brazil. 13418-900. B.O.Box 9

ABSTRACT

This study characterizes the use of risk monitoring mechanisms by coffee and soybean trading and processing companies. It also investigates the role these mechanisms play in the mitigation of certain transaction costs associated to bounded rationality, information asymmetry, and business opportunism in negotiations involving derivatives. The results presented are based on literature and original research, which consisted of interviews with 19 coffee and soybean trading and processing company agents that deal with the management and execution of derivatives trades. The interviews suggest that the interest in formal risk monitoring mechanisms depends strongly on the organizational structure of the business. In family businesses, in which the owner participates actively in negotiations, the interest is limited. In non-family businesses, where there is a clear separation between supervisory upper-management and agents involved with trading, the interest in risk monitoring systems is more evident. This result seems to indicate that the mitigation of transaction costs associated with information asymmetry and opportunism might be stronger motivations for interest in risk monitoring mechanisms than the costs resulting from bounded rationality.

Key-words: transaction cost, financial risk, coffee and soybeans

1. Introduction

Brazilian coffee and soybean processing and trading are responsible for a large part of the country’s agricultural export revenues: Currently, Brazil’s soybean exports are the second largest in the world, and its coffee bean production and exports lead the world. Trading companies and processors often use derivative markets to hedge positions taken in commodities markets. These hedging operations, in spite of allowing reduced risk from negative price movements, expose organizations to the so-called "financial risk" that originates from the unpredictability of economic systems.

There are evidences that these organizations encounter transaction costs related to operation with derivatives. These transaction costs are generated by behavioral postulates inherent to the economic agents, such as bounded rationality and opportunism, which, in turn, can be motivated/enhanced by environmental conditions, such as information asymmetry. The use of formal risk monitoring mechanisms, a practice suggested by many authors, can assist in the economic mitigation of some of these transactional costs.

The present article is intended to characterize the use of risk monitoring mechanisms by coffee and soybean trading and processing companies, as well as the role these mechanisms have in the mitigation of transaction costs generated by company derivatives trading. Important parts of our results are based in original research. A large part of this research was in the form of interviews with a sample of 19 professionals, mostly from coffee trading and soybean processing companies, who are employed to deal with risk management and derivatives trading.

This study has five sections, the first being the introduction. In the second section, we discuss the theoretical framework and present brief summaries of several authors’ understandings regarding the link between transaction costs and operations with derivatives. In addition, the second section contains a discussion of the advantages and disadvantages of current risk monitoring mechanisms, such as the VaR (Value at Risk) risk measure. The third section presents the methodology underlying our empirical research. This methodology depends on interviews with professionals involved, direct or indirectly, with financial operations in the companies under analyses. The fourth section contains the main results obtained from our empirical research, and the last section is this study’s conclusion.

2. Transaction Cost and Financial Risk Management

A large number of empirical studies of the decision process, in fields such as psychology and business administration—summarized in studies by Tversky & Kahneman (1974), Hammond et al. (1998) and in the references cited within these work, support the existence of important transaction costs that are incurred when businesses face complex situations involving risk. These costs arise from an incomplete understanding of the situation and communication problems, two difficulties that characterize aspects of bounded rationality.

The implications of bounded rationality on businesses have been previously suggested and analyzed by Simon (1957) and, more recently, made popular by Williamson (1975). Bounded rationality, a non-intentional characteristic of human behavior, acting in an environment of asymmetric information between agents can lead to intentional opportunistic behavior, which can cause business losses. These notions are summarized by Williamson (1975) and by Milgrom & Roberts (1992) in a more systematic formal discussion of agent theory.

From the studies mentioned above, one can conclude that mechanisms for monitoring risk, which use risk measures intelligible to businesses agents, may facilitate the reduction of transaction costs associated to bounded rationality. By alerting management of potential risk, risk monitoring mechanisms would help to reduce information asymmetry, thereby minimizing clearly inappropriate decisions and blatant opportunism. The use of such monitoring mechanisms, in the context mentioned, could expedite the design of control mechanisms and facilitate decision making more consistently compatible with the business’s interests.

The practical importance of the problems described above is illustrated by the case of Barings Bank. In 1995, the British bank Barings suffered losses of close to US$ 1.3 billion in the Japanese currency and stock markets, caused mostly by decisions made by the agent responsible for the bank’s front and back office operations in Asia (Jorion 1997). The losses were conditioned, on one hand, by clearly inadequate decisions made by that sole agent who believed them to be beneficial to the bank and to himself (bounded rationality/opportunism), and, on the other hand, by the bank’s board of directors’ who did not understand the consequences of the agent’s decisions, consequences which resulted in the bank’s expressive losses (asymmetrical information motivated by lack of adequate monitoring).

Another example of the importance of risk monitoring measures, discovered during an interview conduced in connection with the present study, relates to the possibility of uncontrolled agent opportunism creating business losses. In this case, the possibility of individual self-interest creating undesired business results was brought about by the inadequate design of incentives to promote optimal trading operations: traders received commissions from profits but did not always share losses and were not properly monitored in regards to the increment of company exposure to losses arising from their actions. The use of appropriate risk monitoring mechanisms would have led to the design of more appropriate incentive mechanisms.

In the late 1980s, losses suffered by businesses engaged in derivatives trading motivated the search for risk measures and risk monitoring mechanisms that could be more easily understood by investors and staff with a less sophisticated quantitative background. Many of the mechanisms developed since then have been based on the quantification of risk through measures such as Value at Risk—VaR.

Formally, VaR can be understood as an estimated value for losses established in a distribution of expected rates of return from a portfolio of financial assets that is not to be equaled or exceeded with a determined critical probability over a given period of time. In many contexts, as, for example, under the assumption of normality, the information in the VaR is equivalent, from the logical point of view, to risk measures based on variance or on standard deviation. Nevertheless, market agents with no background in statistical methodology seem to understand the risk information contained in the VaR more easily than that presented by more conventional statistical measures. The VaR tries to capture events that occur on the edges of the portfolio’s return distribution.

Manfredo & Leuthold (1998) suggest that the advantages of VaR are associated to the possibility of its being reported in a currency unit while serving as:

1. a tool used to spread information to control personnel, traders, operators, risk managers, directors, shareholders as an aid in their investment decisions;

2. a tool to help business and investment funds determine resource allocation; and

3. a tool for performance evaluation, used to compare results with some pre-established performance parameter, such as a benchmark.

Due to the meaningful volatility of commodity spot market prices and the exchange rates risk of dealings in the international market, the use of such risk measures by agribusinesses can be advisable. Risk measurement and monitoring is called for by the complex derivatives trading frequently employed by these organizations to hedge their investments and to obtain financial gain through arbitrage and speculation.

Danielsson et al. (1998) observe that regulatory considerations also provide a strong incentive for the use of risk monitoring mechanisms by members of the financial system. Implementation of risk monitoring mechanisms, based in risk measures such as the VaR, demands investments in equipment, software, changes in administrative procedures, and training. Without the pressure of legal requirements, the magnitude of this investment leads organizations outside the financial sector to question if these costs are merited by the magnitude of the benefit

As a last comment, the results obtained in the present study and reported in the following sections suggest that the degree of the interest in risk monitoring mechanisms can be strongly affected by the structure of the agribusiness organization.

3. Methodology

The empirical segment of the present study involved the analysis of qualitative primary information obtained from interviews with people involved, direct or indirectly, with financial decision making in the soybean and coffee complexes’ processing and trading sectors. These two types of Brazilian agribusinesses were chosen for study because they have substantial economic importance in Brazil, operate in domestic and international derivatives markets, and, most importantly, show differences in hierarchy and financial management.

To facilitate this presentation, the agribusiness organizations researched have been classified into two stereotypical categories:

1. family businesses under the owners’ financial management (FBUOFM) – the majority of these businesses are capitalized by domestic sources, controlled through inheritance or otherwise family directed, and are financially managed directly by members of the family. Although many of these companies are small, several of the major coffee exporters are quite large, economically important FBUOFMs

2. businesses under hired professional financial management (BUHPFM) – the majority of these companies are big international players dealing in Brazil. This category includes large domestic agribusiness companies, some family-owned, which are financially administered by hired agents who are not the owners.

The relatively small number of accessible businesses and confidentiality issues associated with the required research information greatly restricted the number of people willing to participate in this study.

The comprehensive and open interviews were intended to capture not only experiences of the person interviewed regarding their own companies but also their knowledge about similar organizations. These interviews were held in person or via e-mail. The personal interviews involved 2 to 3 hour contacts, which were recorded for accuracy. The e-mail interviews were initiated through telephone contact and conducted over several e-mail exchanges.

Interviewees were asked about (a) the processes that lead to financial decisions; (b) the relationship among agents within their business’s several hierarchical levels; (c) existing interest in operations that might reduce the impact of financial risk, such as hedging; and, (d) their perception of the advantages and disadvantages of risk monitoring mechanisms, essentially VaR. For the most part, the questions were intended to characterize the process that results in operations/transactions involving hedges, speculation, and arbitrage in the spot and futures markets. The professionals interviewed have been listed according to their position and the scope of their business, as shown in Table 2.

4. Results and Discussion

Analysis of the material collected in the interviews suggests that a company’s use of more formal risk monitoring mechanisms is determined by its dimension and the amount of direct owner participation in the execution of operations.

In FBUOFMs, more typical of coffee bean export companies, the owners actively participate in operations, acting as traders or directly monitoring other agents’ performance at the operations desk. In these companies, employed traders usually have little autonomy, reporting their decisions directly to company owners. FBUOFMs do not use formal risk monitoring processes and/or specialized agents to supervise these processes. The proximity of company owners is felt to reduce the need for the strict gathering, processing, and analyses of information that characterizes the executed operations and the risks derived from these operations.

The important decisions regarding Brazilian FBUOFM financial operations are made directly by the business’s owners, generally using intuition and practical rules largely inferred from experience. These owners are quite afraid of introducing methodologies they perceive as costly, whose benefits are not clear, and that are dependent on unfamiliar concepts. As the owners are involved in all important decisions, the problem of agent (trader) opportunism is relatively controlled. Decision making through intuitively learned methods can, nevertheless, generate losses associated due to the agents’ bounded rationality. Decisions based on intuition and practical rules (heuristic) can be useful in a general way but result in severe and systematic mistakes in many cases, especially in an environment strongly characterized by complexity and uncertainty (Tversky & Kahneman 1974 and Hammond et al. 1998).

The BUHPFMs examined in our study are typically soybean processors and soybean trading organizations. The salient feature that distinguishes BUHPFMs from FBUOFMs is the clear separation of owners from company operations, monitoring, and decision making. BUHPFM owners are removed from company administrators, the latter being hired professionals. Job positions in BUHPFMs are filled by specialized agents, and job responsibilities are more clearly defined than in FBUOFMs. BUHPFM operations and their monitoring are under the charge of different agents who report to the company board of directors and owners through management.

In FBUOFMs, the relationship between owners and traders is direct and informal: there is no go-between. Decisions regarding financial operations and their monitoring are broadly discussed, although the final decision is the owner’s. Thus, the connection between the board of directors and traders is close and operations often conducted jointly; in many cases the owner working directly with the brokers.

In BUHPFMs, the relationship between the board and traders is intermediated by the back office, which has the risk manager as its main agent. These businesses try to formalize the need for hedging, the limits of exposure, and the danger associated with derivative contracts in reports. These reports are periodically presented to the board of directors and the operations desk by the back office staff. Respecting their superior’s decisions, traders strive to meet specified goals while adhering to defined stop loss parameters, even when they disagree with those decisions. This is not to say that company boards of directors reject the idea of trader speculation. Most boards accept and even encourage their traders to speculate knowing that monitored speculation by informed traders can provide valuable experience and ultimately is good for the company.

The existing organizational structure of the two types of businesses described above is summarized in Figure 1. The Figure diagrams the relationships among different hierarchical levels within FBUOFMs and BUHPFMs. Table 3 presents a synthesis of this information gathered through this study’s empirical research.

Figure 1
- Click to enlarge

Interest in the use of formal methods to measure and report financial exposure from operations is much more evident in BUHPFMs than in FBUOFMs. The use of formal methodologies to monitor and measure risk was found useful to both communicate business exposure to losses at several decision levels and to support the design of mechanisms that motivate traders to make decisions more aligned with organization goals. In businesses where formal risk monitoring mechanisms are used, traders whose operations may cause an exposure to risk above established levels can be chastised and even dismissed. These businesses found that the frequent need to generate the operations reports that formal risk monitoring demands acted as a control mechanism in itself, minimizing potential problems arising from bounded rationality.

5. Conclusion

The BUHPFMs described in the previous section showed strong interest in the improvement of risk measurement and inter-company communication. This is most likely motivated by the significant potential for information asymmetry, a result of their hierarchic structure that clearly separates upper administration from derivatives operations. Such a situation motivates these organizations to seek more efficient mechanisms to facilitate the rationalization of information flow, despite the related costs. Risk monitoring mechanisms based on VaR were perceived as a useful auxiliary tool to insure the proper relationship between economic agents involved in the business’s decision phases and agents working in the business’s financial operations and execution phases.

In FBUOFMs, the closeness of owners to operations tends to minimize the formal flow of risk monitoring information. Unfortunately, this can lead to potential business failures brought about by the agents’ bounded rationality and the use of intuition rather than data as the main guide for decision-making [Hammond et al. (1998) provide a more detailed description of the problems caused by the dependence on intuition for business success].

The potential for transaction costs resulting from the behavioral postulates and environmental conditions previously discussed in this study is summarized in Table 4.

Finally, the difference in interest in risk monitoring mechanisms demonstrated by FBUOFMs and BUHPFMs suggests that the mitigation of transaction costs associated to information asymmetry/opportunism is a stronger motivation for interest in these mechanisms than the costs resulting from bounded rationality.

• DANIELSSON, J.; DE VRIES, C. G.; HARTMANN, P. The cost of conservatism: extreme returns, value-at-risk, and the Basle multiplication factor. Londres: London School of Economics, Jan. 1998. 10p. (Working Paper)
• HAMMOND, J. S.; KEENEY, R. L.; RAIFFA, H. The hidden traps in decision making. Harvard Business Review, p.3-9, Sep./Oct. 1998.
• JORION, P. Value at risk: the new benchmark for controlling derivatives risk. Chicago: Irwin Professional Publishing, 1997. 332p.
• MANFREDO, M. R.; LEUTHOLD, R. M. Agricultural application of value at risk analysis: a perspective. Urbana-Champaign: University of Illinois, 1998. 14p. (Working Paper)
• MILGROM, P.; ROBERTS, J. Economics organization and management. New Jersey: Prentice Hall, 1992. 621p.
• SIMON, H. Models of man: social and rational. New York: John Wiley & Sons, 1957. 279p.
• TVERSKY, A.; KAHNEMAN, D. Judgment under uncertainty: heuristics and biases. Science, v.185, p.1124-1131, 1974.
• WILLIAMSON, O. Markets and hierarchies: analysis and anti-trust implications. New York: The Free Press, 1975. 286p.

Publication Dates

• Publication in this collection
  15 July 2003
• Date of issue
  Mar 2003