swbc024.tmp Journal of Agricultural and Applied Economics, 28,1 (July 1996): 180-192 @ 1996 Southern Agricultural Economics Association A Decision Support Aid for Beef Cattle Investment Using Expert Systems Lawrence L. Falconer, Charles R. Long, and James M. McGrann ABSTRACT The beef cattle investment decision provides an excellent opportunity to increase the economic effi- ciency of beef cattle production. The investment questions that face beef cattle producers are of interest to beef cattle producers, educators, and financial institutions involved in lending to beef cattle producing firms. This study develops a decision support aid utilizing expert system technology to assist beef cattle producers in making well-founded investment decisions with respect to the firm’s beef cattle herd. Key Words: beef cattle investment, decision support, expert systems. The development of sound tools that use economic theory to assist beef producers in making breeding cow investment decisions is an important effort. A major study focusing on competitive issues in the beef sector for the 1990s, conducted through the Hubert H. Humphrey Institute of Public Affairs, pointed to this need (Johnson et al.). One conclu- sion reached in that study was that beef producers must lower their costs of beef production to prevent further loss of market share to competing meats. The report concluded: Two factors are likely to be important in lowering cost of production in the future. The first is the need to use the most efficient production technol- ogy available. The second is the need to consoli- date production into even larger units so that all economies of size are realized. The fact that con- solidation into larger units has been taking place at such a rapid rate in recent years suggests that Lawrence Falconer is an assistant professor with the Texas Agricultural Extension Service, Corpus Christi; Charles Long and James McGrann are professors with, respectively, the Texas Agricultural Experiment Station, Overton, and the Texas Agricultural Extension Service, College Station. there are real economies of size at the production level (p. ii). If the concentration of beef production into larger units in order to lower costs of production becomes more important to the ranching firm’s survival in the future, then sound investment decision making by ranchers regarding beef breeding cattle will be key to their economic survival. Initial results of an analysis of data gathered in the National Cattlemen’s Association’s national da- tabase of integrated resource management stan- dardized performance analysis (SPA) for cow-calf producers (McGrann et al. 1992) clearly show that large operations over the 1990 to 1991 period have had lower costs of production vis-?i-vis medium and small sized producers. Pre-tax cost of produc- tion was found to range from an average of $78.36 per cwt for firms with less than 200 head of cows, to $66.81 per cwt for firms with 200–500 head, to $60.34 per cwt for firms with over 500 head. (For discussion of SPA guidelines, refer to McGrann.) Investment analysis related to breeding cows differs from the analysis of machinery or land in- vestment in the uncertainty associated with the bio- logical aspects of beef cow reproduction and mor- Falcone~ Long, and McGrann: Decision Support Aid for Beef Cattle Investment 181 tality. The investment in cows relative to all other investments made by the beef cattle firm is signifi- cant, increasing the importance of the investment decision in beef cattle. SPA data reflect an average total investment of $2,097 per breeding cow on a cost basis. Thus, assuming a value of $700 to $1,000 per COW,investment in breeding cows would range from one-third to one-half of total beef cow herd investment. The beef cattle investment decision provides an excellent opportunity to increase the economic ef- ficiency of beef cattle production. Investment ques- tions that face beef cattle producers—such as whether to raise or purchase replacement cows, whether to expand the herd size through raising or purchasing cows, how to cull cows with regard to expected reproductive performance, and what breed type of cattle to invest in—are of concern to beef cow-calf producers, educators, and financial institutions involved in lending to beef cow-calf producing firms. The primary objective of this study is to develop a decision support aid (DSA) utilizing expert sys- tem technology to assist beef cattle producers in making well-founded investment decisions with re- spect to the firm’s beef cattle herd. The DSA will be used to analyze how the proposed expansion/ contraction in beef cow herd influences a represen- tative firm’s financial condition and performance. Previous Research The asset replacement problem has received atten- tion in the agricultural economics literature for over 30 years. In his 1965 seminal study, Burt developed a model for economic analysis of asset life under conditions in which there was a chance of failure or 10SS, with replacement falling into planned and random categories, In a subsequent study, Perrin in- troduced a general replacement/investment prin- ciple applicable to both appreciating and depreciat- ing assets, and considered theoretical implications of changing discount rates and market prices (sal- vage values). The specific problem of beef cow replacement decisions was initially addressed in the literature when Rogers carried out an empirical investigation into the beef cow replacement decision under con- ditions of certainty. Bentley, Waters, and Shumway expanded the literature with an empirical study of the problem of determining the optimal replace- ment age for beef cows given stochastic elements relating to the probability of producing a market- able calf and the probability that a cow might die in a particular period. Kay and Rister developed two models to examine the effects of income tax policy on the decision to buy or raise replacement beef cows with a fixed herd size. Innes and Carman ana- lyzed the effects of the Tax Reform Act of 1986 on the decision to either raise or purchase beef cow re- placements. The impact of market price uncertainty on the beef cow replacement decision and herd size began to be addressed in the 1980s. Yager, Greer, and Burt developed optimal policies for marketing cull cows through the use of discrete stochastic program- ming, resulting in changes in cow salvage prices, Bentley and Shumway examined the planning of cattle herd size over cycles in beef cattle inventories and prices. Trapp developed investment principles that resulted in separation of the investment and disinvestment decisions that allow for firm expan- sion or contraction through unequai rates of culling and additions. More recent work has focused on the effects of herd management strategies on the beef cow re- placement decision. The replacement decision was examined by Tronstad and Gum with a stochastic dynamic programming model that took into ac- count herd productivity under multiple calving sea- sons and market price uncertainty. A model was de- veloped by Frasier and Pfeiffer that incorporated the effects of different feeding regimes on expected herd productivity. The review of previous research suggests that many factors enter into the beef cow investment de- cision process, most of which contribute to the level of economic efficiency of the beef cow operation. As noted by Long, Cartwright, and Fitzhugh, “The net efficiency of different systems of beef produc- tion is a function of the genetic and environmental inputs and their interaction” (p. 409). The traditional measures of environmental in- puts would include such factors as availability of grazing or raised feed resources. A broader defini- tion of environmental inputs should include the financial position and performance of the firm, which measures the availability and application of 182 Journal of Agricultural and Applied Economics, July 1996 financial resources. However, the previous works do not explicitly examine the relationship of pro- posed investment decisions to the financial perfor- mance and condition of the beef cow-calf produc- ing firm. Further, previous research has not focused on the effect of proposed investment in beef cows on the herd age composition and the resulting im- pact on the overall financial structure of the firm. In the following section, we present the development of a model that will take these and other factors into account when analyzing the impacts of beef cow investment decisions on the firm’s financial condi- tion and performance. Beef Cow Investment Analysis System In this study, we develop a computerized simulation model of the beef cow firm, identified as the Beef Cow Investment Analysis System (BCIAS). This system is primarily a decision support aid (DSA). Decision support implies the use of computers: (a) to assist managers in their decision-making pro- cesses in semistructured tasks; (b) to support, rather than replace, managerial judgment; and (c) to im- prove the effectiveness of decision making rather than its efficiency (Keen and Morton). The DSA’S impact is on decisions in which there is sufficient structure for computer and analytic aids to be of value, but where the manager’s judgment is essen- tial. Under the manager’s control, the DSA should be a supportive tool, i.e., the DSA is not designed to automate the decision process, predefine objec- tives, or impose solutions (Keen and Morton). The BCIAS simulation is a mathematical model of an actual ranching system. A system here can be defined as a collection of entities (such as cows, calves, grazing resources, or financial resources) that act and interact together to accomplish an end—which, in the case of the BCIAS, is improve- ment in the financial position and performance of the firm. The BCIAS employs discrete-event simu- lation, which models a system as it evolves over time with state variables that change only at count- able intervals (Law and Kelton). The BCIAS contains several features designed to address shortcomings of previous beef cow in- vestment models. These features include: (a) intro- duction of stochastic elements, such as calving per- centage, calf death loss, and weaning weights, into the model to account for uncertainty relating to re- production and production parameters; (b) an ex- pert system analysis of the financial position and performance of the firm under baseline and alterna- tive beef cow investment scenarios; (c) decision support for the user regarding output pricing, pro- duction, and reproduction parameters; and (d) in- corporation of standardized production, reproduc- tion, grazing, and raised feed parameter definitions. The BCIAS is comprised of five modules: the beef cow herd inventory (BCHI) module, the beef cow herd resource inventory (BCHRI) module, the beef cow herd cost/price expectations (BCHCPE) module, the beef cow herd investment analysis (BCHIA) module, and the Agricultural Financial Analysis Expert Systems (AFAES). Figure 1 illus- trates how these modules relate to each other. The data processed in the BCHI, BCHRI, and BCHCPE modules are used in the BCHIA module to simulate the projected performance of both the current cow herd and the proposed alternative cow herd over the user-specified planning horizon. This analysis includes examination of the firm’s financial position and performance for both the current herd and the proposed investment, in addition to an eco- nomic analysis of the proposed investment. The BCHI Module The beef cow herd inventory (BCHI) module is de- signed to provide a vehicle for entering reproduc- tive, productive, and financial data that are related to the firm’s current beef cow herd. The measures for production and reproduction efficiency parame- ters used in the BCHI module are based on SPA recommendations (refer to McGrann). The SPA guidelines designate three primary measures for re- productive performance and two primary measures for productive performance within a herd. The pri- mary reproductive performance measures are (a) calving percentage, (b) calf death loss, and (c) calf crop; the primary measures for productive perfor- mance are (a) actual weaning weights, and (b) pounds weaned per exposed female. The production and reproduction measures were used to develop default production parameter estimates that are contained in the BCHI module. These default parameters are intended as examples for users who have the capability to generate such values from their own records, in which case these parameters may be overridden. However, in the Falconer Long, and McGrann: Decision Support Aid for Beef Cattle Investment Beef Cow Herd Inventory Module (BCHI) E L 183 Beef Cow Herd Cost/Price Expectations Module (BCHCPE) Investment Analysis I Module (BCHIA) * Figure 1. Beef Cow Investment Analysis System (BCIAS) modules case where production records are missing, the de- fault parameters can provide a basis for beginning the analysis. The data for estimating default production and reproduction parameters were obtained from a study that included 5,903 calving records for 988 cows at the Texas A&M Research Center at McGregor, Texas. (See Long et al, for details of early management techniques applied to these cattle.) Cows were managed so as to conceive whenever they were physiologically able; cows fail- ing to conceive within 18 months of their last partu- rition date and found to be open were culled. When necessary, cows were also culled for physical un- soundness (McElhenney et al,), Following the work of Greer, Whitman, and Woodward, the Bernoulli distribution was esti- mated by breed type for the calving percentage re- production parameter. The Bernoulli distribution for a random variable x is shown in equation ( 1): (1) f(x; f)) = 6P(1 – 0)’-’, forx=O, l. For this study, the definition for success will be that the cow actually calved within any given pe- riod of time. With the additional assumption of in- dependence between stages of the trials, the Ber- noulli distribution can be extended to become the binomial distribution (Freund and Walpole). Here, we assume that at any specific age, the probability of any cow calving does not depend on any other cow calving that is the same age. The binomial dis- tribution for a random variable x is shown in equa- tion (2): (2) f(x; n, 6) = () n W(1 – 6)’-’, x forx=O, l n.,. ..> For simulation purposes, the number of cows that are available to be exposed would be specified as the number of trials, n, with x being the total num- ber of cows calving. The binomial distribution is used to estimate the calf crop reproduction parameter from the data by breed and age group. The probability of success is defined as a calf born. The simulated number of successes (here, live calves born) is adjusted for death loss and then divided by the simulated num- ber of cows exposed in the corresponding breeding season to arrive at the simulated calf crop or wean- ing percentage. Calf death loss and calf crop reproduction per- formance measures are based on the number of calves born option in the SPA guidelines. Breed type has been shown to be a significant source of 184 Journal of Agricultural and Applied Economics, July 1996 Table 1. Calf Death Loss Percentage and Weaning Weights by Breed Types Percent Mean Weaning Breed Death Loss Weight (lbs.) Angus Brahman Hereford Holstein Jersey Angus X Brahman Angus X Hereford Angus X Holstein Angus X Jersey Brahman X Hereford Brahman X Holstein Brahman X Jersey Hereford X Holstein Hereford X Jersey Holstein X Jersey 4.34 10.58 5.18 4.76 8.21 5.96 7.19 5.81 6.05 4.59 9.34 5.05 3.84 6.29 9.04 426 435 396 503 409 456 435 481 440 457 515 458 490 438 467 Source: Falconer variation for calf survival from birth to weaning (McElhenney et al.). The estimated default calf death loss parameters by breed type are presented in table 1. To arrive at the SPA reproduction effi- ciency measure of calf crop for projection pur- poses, the number of calves born is calculated from starting exposed cow numbers and default calving percentages, to arrive at total calves born. The total number of calves born is then adjusted for calf death loss to arrive at total calves weaned. The SPA primary reproduction efficiency measure for calf crop is calculated by dividing total calves weaned by the total number of exposed cows, The SPA productive performance measure for actual weaning weights is estimated for each cow breed type included in the study conducted by Long et al, The distributions used in the simulation process are based on normal curves, with estimated parameters shown in table 1, The average weaning weight is then simulated as a normal distribution that uses the mean and standard deviation of the samples by breed type, with a user-specified adjust- ment available to be used to correct for age of dam effects. The weights are then multiplied by the calves weaned in the respective breed type and age group, and then summed to arrive at the total pounds of calves weaned. The simulated SPA per- formance measure of pounds weaned per exposed cow is calculated by dividing the total pounds weaned by the number of exposed cows. While cull cow sales is not a primary perfor- mance indicator designated by the SPA guidelines, it is addressed within the guidelines. To estimate weight of cull cows for sale, we drew from the work of Nelsen, Long, and Cartwright. The estimation procedure used here is shown by equation (3) (Brody): (3) Y, = A – Be-”, where Y,is the weight of the animal in kilograms at time t,A is the asymptotic weight, B is a constant of integration, k is the measure of the rate at which the curve is approaching the asymptote, and t is the time in months. This model for weight at any given time in months for breed type and condition score is used in the BCHIA module to calculate sale weight of cull cows by breed type and age group. (See Nelsen, Long, and Cartwright for a more de- tailed discussion of this method.) The BCHRI Module To evaluate the proposed beef cow investment, the resources available for input into the beef cow-calf enterprise must be accurately inventoried. This is accomplished through the beef cow herd resource inventory (BCHRI) module. To examine the perfor- mance of grazing and raised feed resources in the beef cow-calf enterprise, the BCIAS model incor- porates the following two primary measures as de- fined by the SPA guidelines: (a) grazing acres per exposed female, and (b) pounds weaned per acre utilized by the cow-calf enterprise. The simulation model calculates the SPA grazing and raised feed performance grazing acres per exposed female measurement by dividing the number of grazing acres by the simulated number of exposed females. To facilitate the establishment of baseline mea- sures of the firm’s financial condition and perfor- mance and to provide a general format for analysis of projected results from investment in beef cattle, the BCHRI module uses the FINYEAR software package (McGrann, Parker, and Neibergs). FIN- YEAR is a computer program that was created to assist in the development of a set of financial state- ments for a single operating year. The program’s formulation has been closely coordinated with the FalconeC Long, and McGrann: Decision Support Aid for Beef Cattle Investment 185 published Recommendations of the Farm Financial Standards Task Force (Farm Financial Standards Task Force). The task force has developed financial analysis standards and terminology for agricultural businesses. The BCHCPE Module The beef cow herd costJprice expectations (BCHCPE) module is designed to aid the user in the development of cost expectations for grazing and forage production, discount rates used in eco- nomic investment analysis, and output prices for the beef cow enterprise. The output price expecta- tions specifications for the beef cow enterprise used in this study are grouped into three major catego- ries: (a) expectations based on past own informa- tion, (b) subjective expectations of the model user, and (c) expectations based on solutions from a structural agricultural sector model, A naive expec- tations model is used to represent the first category. For subjective price expectations, projected mean cattle prices are elicited from the producer, and then random deviates of historical prices are generated to create the price probability distribution (Rich- ardson et al,). Expected cattle prices for the third category are taken from the AG-GEM model (Pen- son and Taylor), a structural econometric model that specifically links the agricultural sector with the general economy. The BCHIA Module To carry out an economic analysis of the proposed investment for the user-specified planning horizon, the beef cow herd investment analysis (BCHIA) module applies a net present value investment model (as suggested by Barry, Hopkin, and Baker), modified to include a time variant discount rate [equation (4)]: “) “v=-’N”+%++,,,I+[UH where NPV is the net present value, INV is the ini- tial investment, P. is the net cash flows attributed to the investment that can be withdrawn in period n, V~ is the terminal investment value, N is the length of the planning horizon, and i,, is the discount rate. For a single proposed investment, the decision rule for the economic analysis would be to accept the investment on an economic basis if the NPV is greater than zero. In the BCHIA module, the net cash flows and terminal values include both sto- chastic and deterministic elements. The BCHIA module uses the net cash flow from operation of the cattle enterprise as the income flow per period stream in the net present value model. The simulation results also show the range that the simulation generates for net cash flow from opera- tions, allowing the user to measure the range of out- comes as a factor of risk. Assuming independence of cash flows, the standard deviation of the net pres- ent value can also be calculated as shown in equa- tion (5) and used as a measure of risk (Bussey): where V is the variance of the net present value, var, is the variance of the cash flow in time period t,and i, is the discount rate by period, BCHIA Module Output as Input for AFAES Program The output from the BCHIA module is used as in- put into the Agricultural Financial Analysis Expert Systems (AFAES) program. To analyze the firm’s projected performance, the AFAES projected op- erating year performance expert system component uses four years of historic balance sheets, three years of income statements, and three years of statements of cash flow data, along with projected balance sheet, projected income statement, and projected cash flow data. Through this AFAES pro- gram component, a summary report is provided of actual historic and projected measures that are ex- amined, as well as a graphic overall diagnostic analysis of the firm’s financial position and perfor- mance. (For a comprehensive discussion of the AFAES program, see McGrann et al. 1990.) Base Simulation Data To validate the BCIAS, we selected an investment problem facing the Texas A&M University (TAMU) Farm, located in the Brazes River Valley of central Texas. The simulations that were carried out examine two possible courses of action: (a) 186 Journal of Agricultural and Applied Economics, July 1996 Table 2. Probability of Angus X Brahman and Brahman X Hereford Cows Calving Whhin 365 Days of Previous Parturition Age in Years Breed 3 4 5 6 7 8 9 10 11 12 Angus x Brahman 0.086 0.897 0.811 0.833 0.919 0.750 0.704 0.833 0.800 0.500 Brahman X Hereford 0.172 0.797 0.830 0.875 0.878 0.714 0.800 0.682 0.733 0.364 Source: Falconer maintaining the cow herd at present levels, or (b) liquidating the cow-calf enterprise and leasing out the land on which the enterprise was operating. Us- ing the three methods of developing expectations (previously discussed), the BCIAS was run to eval- uate differences in economic analysis results under these options. Results from the simulation using the AG-GEM structural model expectations for the continuation and disinvestment strategies were then processed through the AFAES program. The data used as a basis for the simulations were taken from actual SPA reports for the TAMU Farm. The farm’s cattle are comprised of two main breed types—Beefmaster and Brangus. Thus, two sets of calving probabilities were generated. Be- cause of the breed composition of Beefmaster cattle, which includes Brahman, Hereford, and Shorthorn, the Brahman X Hereford data set from the McGregor Experiment Station was selected to represent the Beefmaster cows, The Brangus were likewise represented by the Angus X Brahman data set from the McGregor Experiment Station because of the Brahman and Angus composition of the Brangus breed. The McGregor data were deemed appropriate to represent the calving performance of the TAMU Farm cows due to the proximity of the TAMU Farm to the McGregor, Texas, area and the overall management practices applied to the ani- mals. These practices included adequate levels of nutrition and the use of artificial insemination, which can substantially improve reproductive per- formance. To generate the required Bernoulli probabilities used in the binomial simulation for the model, a transition matrix moving from parturition probabil- ity data to a probability of parturition by age group under the desired culling strategy was developed for both breed types. The transition probabilities for Angus X Brahman and Brahman X Hereford cows are presented in table 2. The TAMU Farm currently devotes 748 total acres to the beef cow enterprise. The grazing re- sources for the farm include 643 acres of improved perennial pasture, in addition to five acres of annual pastures or forage crops. The farm’s raised feed re- sources include 100 acres of improved perennial grasses devoted to hay production. During fiscal year 1991, the TAMU Farm incurred $102,372 in total direct cash expenses relating to the beef cow- calf enterprise, $2,016 in direct noncash expenses relating to the beef cow-calf enterprise, $3,229 in total indirect cash expenses, and $1,245 in total in- direct noncash expenses. This cost structure will serve as the base for cost structure estimates for all simulations in this study, Table 3 shows the base cattle prices used in the simulation runs, The base for the naive expecta- tions simulation for calf prices was the weighted average actual price received for all calves at the TAMU Farm for 1991, which was $83.85 per cwt. The AG-GEM price expectations simulation for calf prices was generated by using the annual per- centage change in AG-GEM forecasted calf prices applied to the $83.85 per cwt base. Calf prices for the subjective expectations simulation were taken from unpublished survey data (Falconer and Neib- ergs). Highest and lowest expected prices, along with expected prices, were gathered from over 60 cow-calf producers in the spring of 1991 for the 199 1–95 period. The mean of the survey data was used as the parameter input into the subjective price distribution. As with calf prices, the base for the naive expec- tations simulation for cow prices was the weighted average actual price received for all cows at the TAMU Farm for 1991, which was $62.92 per cwt. Falcone~ Long, and McGrann: Decision Support Aid for Beef Ca~tleInvestment 187 Table 3. Base Cattle Prices Used in Simulation Runs Year Simulations 1 2 3 4 5 ---- Calf Prices ($/cwt) ----------- --- Naive Expectations 83.85 83.85 83.85 83.85 83.85 AG-GEM Model 87.20 86.33 86.59 87.02 87.46 Subjective Expectations 93.21 90.21 88.21 87.21 87.21 ---- - Cull Cow Prices ($/cwt) - - - - - - - - - - - Naive Expectations 62.92 62.92 62.92 62.92 62.92 AG-GEM Model 67.29 65.62 64.82 64.06 63.31 Subjective Expectations 55.21 52.21 50.21 49.21 49.21 Table 4. Projected Cow Herd Composition and Production by Year cows Weaned Calf Cull cull Exposed Birth Production Cows Sales Year (head) (head) (lbs.) (head) (lbs.) 1 158 117 51,363 28 32,558 2 158 114 50,052 32 37,204 3 159 111 48,736 36 41,745 4 159 117 51,374 20 23,191 5 159 110 48,298 23 22,008 The AG-GEM price expectations simulation for cow prices was generated by using the annual per- centage change in AG-GEM forecasted cow prices applied to the $62.92 per cwt base. The cow prices for the subjective expectations simulation were generated by adjusting the survey data taken from Falconer and Neibergs for a constant basis. The ba- sis between cow and calf prices was assumed to be $38 per cwt (table 3). The initial composition of the TAMU Farm herd by age and breed type was used as the basis for sim- ulation runs. The culling strategy imposed on the herd was to cull any cow over three years of age that did not calve, and to cull all cows after reaching 11 years of age. The heifer must calve to enter the herd at all, All replacement animals were assumed to be retained from the Beefmaster herd. Table 4 presents the projected composition and production of the herd by year. The projected herd becomes younger under the specified culling and replacement strategy (figure 2), In the initial cattle inventory, 39% of the total cows are less than five years of age, while in the fifth year of the simulation, 69% are less than five years of age. The number of cows under five years of age peaks at 72% in the fourth simulated year. The herd becomes less productive over time (table 5), dropping from 325 pounds weaned per cow ex- posed in the first year of the simulation to 304 pounds in the final year of the five-year planning horizon. This productivity decline is due to lower reproductive performance of the younger cows, pri- marily cows that are calved as two-year-olds, and attempts to rebreed and calve at three years of age. The net farm/ranch income from operation of the beef cattle enterprise is literally the bottom line for the income statement in the BCHIA module. The net income is calculated in this model by sub- tracting accrued direct and indirect expenses from the accrued gross revenue, and then adjusting for total interest expense. The projected accrual- adjusted income statement for the beef cow enter- prise utilizing AG-GEM cost and price expecta- tions is shown in table 6. Investment Analysis Results Simulation results for net farm income from opera- tion of the cattle enterprise were generated for each of the three alternative cattle price expectations methods used in the BCHIA module for the final investment analysis. The AG-GEM forecasted in- terest rate for non-real estate loans was used for the investment analysis in each scenario, The interest rate on non-real estate loans was selected for use as the discount rate because the TAMU Farm does have a small amount of debt which could be paid 188 Journal of Agricultural and Applied Economics, July 1996 Table 5. Standardized Performance Measures for Simulated Herd Meas. Year Unit 1 2 3 4 5 Calving percent % 74.1 72.2 69.8 73.6 69.2 Percent calf crop % 70.9 69.6 67.3 70.4 66.7 Average weaning weight Ibs. 459 455 455 459 456 Pounds weaned per exposed female lbs. 325 317 307 323 304 Total acres per exposed female acres 4.7 4.7 4.7 4,7 4.7 Pounds weaned per acre utilized by cow-calf enterprise lbs. 68.7 66.9 65.2 68,7 64.6 Financial cost (noncalf revenue adjusted) per cwt $ I .94 2.08 2,22 2.47 2.81 120- 100 80 60 40 20 0 ,----- r , .– – --T---- I 1 2 3 4 5 Year Figure 2. Age composition of projected herd down (as an alternative to keeping capital tied up in the cattle operation). The results for net farm income from operation of the cattle enterprise were not encouraging for any of the simulations, and in particular for the AG- GEM expectations simulations. The sharp increase in predicted feed costs early in the planning hori- zon causes a large increase in the total cost struc- ture over the AC-GEM expectations planning sce- nario. This cost increase, coupled with steady to declining cattle prices over the AC-GEM expecta- tions planning scenario, leads to a negative net cash flow from operation of the beef cattle enterprise at the TAMU Farm. The net present value of the net cash flow from operation of the cattle enterprise combined with the market value of the ending cattle inventory is – $238,253 for the AC-GEM ex- pectations planning scenario, The results for net farm income from operation of the cattle enterprise were also negative over the entire period for the naive expectations simula- tions. However, with costs held level, net farm income from cattle operations is higher in com- parison to the AC-GEM expectations planning sce- nario. The net present value of the net cash flow from operation of the cattle enterprise combined with the market value of the ending cattle inventory is –$269,430 for the naive expectations planning scenario. The standard deviation of the net present value of the net cash flow from operations for the naive expectations option is $4,731, which repre- sents the smallest of the standard deviations calcu- lated for all expectations options. The subjective expectations simulations like- wise yielded negative results over the entire period for net farm income from operation of the cattle en- terprise. Due to the expectation of declining calf and cow prices over much of the planning horizon, the subjective model produced the lowest returns of all three expectations scenarios. Under the subjec- tive expectations planning scenario, the net present value of the net cash flow from operation of the cattle enterprise combined with the market value of the ending cattle inventory is – $275,275. The stan- dard deviation of the net present value of the net cash flow from operations for the subjective price expectations option is $8,664—the largest of all the standard deviations for any expectations option. A result of increased variability when em- ploying the subjective expectations option relative to the naive and AC-GEM expectations options is not surprising. Since the subjective expectations option specifically introduces uncertainty in the Falconec Long, and McGrann: Decision Support Aid for Beef Cattle Investment 189 Table 6. Beef Cow Herd Investment Analysis (BCHIA) Module’s Projected Income Statement for AFAES Evaluation of Continued Cattle Operation Strategy Year 1 2 3 4 5 Gross Revemre Total Direct Cash Expenses Total Direct Noncash Expenses Total Direct Operating Expenses Gross Margin Total Indirect Cash Expenses Total Indirect Noncash Expenses Total Indirect Operating Expenses Income After Indirect Expenses Total Interest Cash Expenses Total Interest Noncash Expenses Total Interest Expenses Total Pre-Tax Farm/Ranch Expenses Net FarrM?anch Income from Operation 61,557 115,120 2,151 117,272 (55,715) 3,229 1,245 4,474 (60, 189) o 0 0 121,746 (60, 189) 47,058 121,957 2,229 124,185 (77,127) 3,229 1,245 4,474 (81,601) o 0 0 128,659 (81,601) .($) -- 59,880 128,248 2,311 130,559 (70,679) 3,229 1,245 4,474 (75,153) o 0 0 135,033 (75,153) ---- -. 62,258 135,149 2,401 137,551 (75,293) 3,229 1,245 4,474 (79,767) o 0 0 142,025 (79,767) 56,010 142,453 2,495 144,948 (88,938) 3,229 1,245 4,474 (93,412) o 0 0 149,422 (93,412) Note: The AC-GEM expectations option is used as a baseline for this strategy. output price mechanism by specifying a price dis- tribution, it follows that the subjective expectations option will lead to larger variances in returns rela- tive to the single-valued estimates employed in the AG-GEM and naive expectations scenarios. AFAES Comparison of Two Projected TAMU Farm Strategies Given the negative results generated in the previous investment analysis, two projected strategies for the TAMU Farm were selected for AFAES evaluation and comparison. The first alternative is one of con- tinued operation using the AG-GEM expectations option as a baseline (table 6). The second option is to sell all the cattle and lease the land currently occupied by use of the cattle (table 7). The land utilized by the cattle operation at the TAMU Farm was deemed to have 300 acres suit- able for growing cotton, and was assumed to have a cash lease rate of $40 per acre. The balance of the land utilized by the farm’s cattle operation (448 acres) was deemed to be suitable only as grazing land and was assumed to have a cash lease rate of $16 per acre. The projected income statement for the sell at the end of 1992 strategy is shown in table 7, For purposes of the expert system analysis, operations other than the cattle enterprise at the TAMU Farm were treated using a naive expec- tations approach. Under the continued-operation strategy, the expert system analysis on the projected financial performance of the farm fell from accept- able to unfavorable after the 1993 operating year. The AFAES diagnosis cited the extremely poor profitability of the operation (rating profitability at –29.6 based on a range of – 30 to +30) and deteri- oration in firm growth as reasons for the change to an unfavorable performance rating. In contrast, the AFAES diagnosis gave the TAMU Farm a favorable rating of 15 (from a range of – 15 to + 15) for its liquidity position, but noted that the liquidity position of the firm was not show- ing improvement over time. Because of a lack of debt, the AFAES diagnosis also gave the farm a fa- vorable rating for debt repayment capacity of 12,5 (from a range of –25 to +25), In addition, the di- agnosis produced a favorable rating of 6.7 (from a range of – 20 to +20) for the farm’s solvency posi- tion, but warned that the solvency position of the firm has been declining. Conversely, the AFAES evaluation of the sell at the end of 1992 strategy gave the TAMU Farm acceptable ratings for the entire planning horizon. For example, the diagnosis assigned a favorable rating of 15 (from a range of – 15 to + 15) for the 190 Journal of Agricultural and Applied Economics, July 1996 Table 7. Beef Cow Herd Investment Analysis (BCHIA) Module’s Projected Income Statement for AFAES Evaluation of Sell at End of 1992 Strategy Year 1 2 3 4 5 Gross Revenue Total Direct Cash Expenses Total Direct Noncash Expenses Total Direct Operating Expenses Gross Margin Total Indirect Cash Expenses Total Indirect Noncash Expenses Total Indirect Operating Expenses Income After Indirect Expenses Total Interest Cash Expenses Total Interest Noncash Expenses Total Interest Expenses Total Pre-Tax Farm/Ranch Expenses Net Farm/Ranch Income from Operation 39,687 102,372 2,016 104,388 (64,701) 3,229 1,245 4,474 (69,175) o 0 0 108,862 (69, 175) 19,168 0 0 0 19,168 3,229 1,245 4,474 14,694 0 0 0 4,474 14,694 ($) - - 19,168 0 0 0 19,168 3,229 1,245 4,474 14,694 0 0 0 4,474 14,694 19,168 0 0 0 19,168 3,229 1,245 4,474 14,694 0 0 0 4,474 14,694 19,168 0 0 0 19,168 3,229 1,245 4,474 14,694 0 0 0 4,474 14,694 farm’s liquidity position, and cited strong improve- ment over time, Solvency, repayment capacity, and growth were also given favorable ratings, although AFAES noted slow increases in earned equity as a reason not to give the highest possible ratings for firm growth. The TAMU Farm’s projected profit- abilityy received an extremely low rating of – 14.47 (from a range of – 30 to +30) under the sell at the end of 1992 strategy. While the farm is profitable under this strategy, the rate of return on farm assets was cited as being low, and not showing any im- provement over time. Model Validation The SPA primary performance measures for the TAMU Farm simulation were compared with the latest SPA summary data (McGrann et al. 1992) for validation purposes (table 5). Although the simu- lated TAMU Farm calving and calf crop percent- ages are below the weighted average of the SPA summary, they are within the SPA summary’s re- ported range, as are the simulated average weaning weights, pounds weaned per exposed female, and acres per exposed female. The simulated TAMU Farm pounds weaned per acre utilized by the cow- calf enterprise are above the weighted average of the SPA summary, and are also within the summa- ry’s reported range.. . However, the TAMU Farm simulation shows the current operation’s projected cost of production to be extremely high relative to the SPA summary financial data. The observed range in the SPA sum- mary for weaned calf cost per cwt reflected a low of$31 per cwt and a high of $141 per cwt. In none of the projected years did the TAMU Farm simula- tion have a weaned calf cost less than the highest observed cost. Care should be taken in interpreting these particular projections, since they are depen- dent upon the changes in costs based on AG-GEM forecasts. Summary and Concluding Remarks In this study, we have developed a decision support aid that examines the impact of proposed invest- ment decisions in beef cattle on the firm’s financial performance and position. The model depends heavily on the use of electronic spreadsheets for generation of projections that are used as inputs in a computerized expert system. Data requirements are specific, i.e., the initial inventories of animals and related resources, as well as financial statement information, are required. However, these informa- tion requirements are not overly burdensome, as shown in the example where baseline data were taken from SPA reports and cattle inventories by age group. Falcone6 Long, and McGrann: Decision Support Aid for Beef Cattle Investment 191 The model is currently being used at the county agricultural extension level. While the main goals of this study have been achieved, further work is being conducted with the model to exploit the capa- bilities of the latest generation of computer soft- ware. These enhancements center on making the data transfer between modules more transparent to the user, which will reduce the training time and effort required to generate the analysis. References Barry, P. J., J. A. Hopkin, and C. B. Baker. Financial Management in Agriculture. Danville IL: The Inter- state Printers and Publishers, Inc., 1983. Bentley, E., and C, R. Shumway. “Adaptive Planning Over the Cattle Price Cycle.” S. J. Ag~ Econ. 13(1981):139-48. Bentley, E., J. R. Waters, and C. R. Shumway. “Determin- ing Optimal Replacement Age of Beef Cows in the Presence of Stochastic Elements.” S. J. Agr Econ. 8(1976):13-17. Brody, S. Bioenergetics and Growth. New York: Rein- hold Publishing Co., 1945. Burt, 0. R. “Optimal Replacement Under Risk.” J Farm ~COIZ. 47( 1965):324-45. Bussey, L. E. The Economics of Industrial Projects. Englewood Cliffs NJ: Prentice-Hall, Inc., 1978. Falconer, L. L. “Beef Cow Enterprise Investment/Disin- vestment Analysis with Expert Systems.” Unpub- lished Ph.D. dissertation, Texas A&M University, December 1992. Falconer, L. L., and J. S. Neibergs. “Replacement Cow Market Development Study.” Unpublished paper, Dept. of Agr. Econ., Texas A&M University, 1991. Farm Financial Standards Task Force. Recommendations of the Farm Financial Standards Task Force. Ameri- can Banker’s Association, Washington DC, May 1991. Frasier, W. M., and G. H. Pfeiffer. “Optimal Replace- ment and Management Policies for Beef Cows.” Ame~ J Ag~ Econ. 76(November 1994):847-58. Freund, J. E., and R. E. Walpole. Mathematical Statis- tics, 4th ed. Englewood Cliffs NJ: Prentice-Hall, Inc., 1987. Greer, R. C., R.W. Whitman, and R. R. Woodward. “Esti- mation of Probability of Beef Cows Being Culled and Calculation of Expected Herd Life.” J Animal Sci. 51(1980): 10-19. Innes, R., and H. Carman. “Tax Reform and Beef Cow Replacement Strategy!’ West J. Agz ECCM. 13(1988):254–66. Johnson, D. G., J. M. Conner, T. Josling, A. Schmitz, and G. E. Schuh. Competitive Issues in the Beef Sector: Can Beef Compete in the 1990s? Humphrey Institute Rep. No. 1, Hubert H. Humphrey Institute of Public Affairs, University of Minnesota, 1989. Kay, R. D., and E. Rister. “Income Tax Effects on Beef Cow Replacement Strategy.” S. 1 Agz Econ. 9(1977):169-72. Keen, P. G. W., and M. S. Morton. Decision Support Sys- tems: An Organizational Perspective. Reading MA: Addison-Wesley Publishing Co., 1978. Law, A. M., and W. D. Kelton. Simulation Modeling and Analysis. New York: McGraw-Hill, Inc., 1982. Long, C. R., T. C. Cartwright, and H. A. Fitzhugh, Jr. “Systems Analysis of Sources of Genetic Variation in Efficiency of Beef Production: Cow Size and Herd Management.” J. Animal Sci. 40(1975):409-20. Long, C. R., T. S. Stewart, T. C. Cartwright, and T. G. Jenkins. “Characterization of Cattle of a Five-Breed Diallel: I. Measures of Size, Condition, and Growth in Bulk” 1 Animal Sci. 49(1 979):41 8–31. McElhenney, W. H., C. R. Long, J. F. Baker, and T. C. Cartwright. “Production Characters of First- Generation Cows of a Five-Breed Diallel: Reproduc- tion of Young Cows and Preweaning Performance of Inter Se Calves.” J. Animal Sci. 61(1985):55-65. McGrann, J. M. “Guidelines for Production and Finan- cial Performance Analysis for the Cow-Calf Pro- ducer.” Prepared for the National Cattlemen’s Associ- ation, through Dept. of Agr. Econ., Texas A&M University, 199 I. McGrann, J. M., K. L. Karkosh, L. L. Falconer, and J. S. Neibergs. “AFAES-Agricultural Financial Analysis Expert Systems.” Texas Agr. Exp. Sta., Dept. of Agr. Econ., Texas A&M University, 1990. McGrann, J. M., J. Parker, L. L. Falconer, G. Clary, J. S. Neibergs, and P. Gutierrez. “Standardized Perfor- mance Analysis (SPA) Shows Low Profit, Effects of Economy of Size and Opportunities for Change in the Cow-Calf Sector.” Texas Agr. Ext. Ser., Dept. of Agr. Econ., Texas A&M University, 1992. McGrann, J. M., J. Parker, and J. S. Neibergs. “FlN- YEAR—Farm/Ranch Financial Statement Prepara- tion Package.” Computer program, Texas Agr. Exp. Sta., Dept. of Agr. Econ., Texas A&M University, 1991. Nelsen, T. C., C. R. Long, and T. C. Cartwright. “Post- inflection Growth in Straightbred and Crossbred Cattle: I. Heterosis for Weight, Height, and Maturing Rate.” J. Animal Sci. 55( 1982):280-92. Penson, J. B., and C. R. Taylor. “Modeling the Interface Between Agriculture and the General Economy.” AFPC Policy Work. Pap. No. 90-13, Agricultural and Food Policy Center, Texas Agr. Exp. Sta./Agr. Ext. Ser., Dept. of Agr. Econ., Texas A&M University, 1990, 192 Journal of Agricultural and Applied Economics, July 1996 Perrin, R, K. “Asset Replacement Principles.” Arne~ J. Ag~ Econ. 54(1972):60-67. Richardson, J. W., P. Zimmel, D. Gerloff, D. Anderson, G. Helms, A. Lovell, A. Gray, G. Allen, M. Cochran, S. Wong, D. Justice, and M. Rister. “CIRMAN: ,Crop Insurance Risk-Management Analyzer.” Agricultural and Food Policy Center, Texas Agr. Exp. Sta./Agr, Ext. Ser., Dept. of Agr. Econ., Texas A&M Univer- sity, 1992. Rogers, L. “Economics of Replacement Rates in Com- mercial Herds.’’ Animal,Scici. 34(1972):921–25. Trapp, J.N, “Investment and Disinvestment Principles with Nonconstant Prices and Varying Firm Size Ap- plied to Beef-Breeding Herds.’’ Ame~ J. Ag~ Econ. 68( 1986):69 1-703. Tronstad R., and R. Gum. “Cow Culling Decisions Adapted for Management with CART.” Ame~ J. Ag~ Econ.76(May 1994):237-49. Yager, W. A., R. C. Greer, and O. R. Burt. “Optimal Poli- cies for Marketing Cull Beef Cows.” Ame~ J. Agx Econ. 62(1980):456-67.