AN INTERACTIVE DECISION MODEL INTEGRATING BROILER PRODUCTION AND PROCESSING RESPONSIVENESS TO CONSUMER AND PRODUCER PRICES

AN INTERACTIVE DECISION MODEL INTEGRATING BROILER PRODUCTION AND PROCESSING RESPONSIVENESS TO CONSUMER AND PRODUCER PRICES ECIO F. COSTA Professor, Departamento de Economia, Universidade Federal de Pernambuco Praça Professor Fleming, 50 / Apto. 1201, Recife, PE, Brazil 52.050-180. ecio@yahoo.com JACK E. HOUSTON Professor, Department of Agricultural and Applied Economics, University of Georgia 301 Conner Hall, Atthens, GA, USA 30602. jhouston@agecon.uge.edu Poster paper prepared for presentation at the International Association of Agricultural Economists Conference, Gold Coast, Australia, August 12-18, 2006 Copyright 2006 by Ecio F. Costa and Jack E. Houston. All rights reserved. Readers may make verbatim copies of this document for non-commercial purposes by any means, provided that this copyright notice appears on all such copies. 1

2 AN INTERACTIVE DECISION MODEL INTEGRATING BROILER PRODUCTION AND PROCESSING RESPONSIVENESS TO CONSUMER AND PRODUCER PRICES INTRODUCTION Seasonal markets for poultry products, especially those for whole broilers for roasting and those for cut-up birds, often motivate decisions on alternative production and processing of birds, the inputs which will be used, and levels and timing of processes. Relative prices of inputs, and, in particular, the main ingredients of feed formulations in the case of broiler production, may also drive changes in production and processing activities. Previous decision models in this industry have treated production and marketing options separately, with production cast in a least-cost framework restricted by predetermined bird size, number of days on feed, and/or protein levels, among others. We develop and evaluate an interactive decision model to analyze broiler production and processing profitability by exploiting alternative feed protein sources from the perspective of a poultry integrator that faces changing prices of inputs (e.g., soybean meal and cottonseed meal) and consumer-driven prices of outputs (whole carcass or cut-up parts of broilers) When searching for alternative protein sources for poultry or livestock feed, in addition to the feeding response, quality and safety aspects, integrators must also be attentive to the availability of those sources, storage and marketing of alternatives. Among all protein sources, after soybean meal (SBM), cottonseed meal (CSM) is the most traded and it is priced in at least seven different major markets in the United States (Feedstuffs, 2000). If the use of CSM or other oil meals, such as peanut, for poultry production became a reality, oilseed producer gains/losses would also be substantively

3 altered in areas which currently produce soybeans, cottonseed, or peanuts. However, profitable use of other oil meals requires that the price must be lower than the price of SBM. Not only that, if not as productive, the price for CSM (or other meals) must be such that it compensates for the lower physical productivity of the CSM-fed broilers. Price differences of protein sources play an important role, because protein sources in broiler diets account for approximately 30% of the total dietary costs for high protein level feeds. Historical data on prices show that SBM price has always been higher than the price of CSM for several markets in the United States (Feedstuffs, 2000). With its high protein concentration, cottonseed meal could be used as a protein source in the feed formulation for poultry production in the United States (US). However, could cottonseed meal or other protein sources efficiently and profitably replace soybean meal, currently the primary protein source used in broiler production in the US? Cottonseed yields 900 pounds (45%) of CSM for each ton that is crushed (National Cottonseed Products Association, NCPA, 1999). While high in protein (usually sold between 40% and 42%), CSM s most limiting amino acid is lysine (Grau, 1946, and Anderson and Warnick, 1966), and its use in the US has been concentrated in feed for livestock (NCPA, 1999). The main obstacle in dealing with CSM for poultry and livestock rations is the presence of gossypol, a natural toxin present in the cotton plant, due to current processing techniques that cause lysine deficiency (Fisher and Quisenberry, 1974). In Canada, however, CSM is used as a protein source for broiler chickens in addition to SBM and other protein sources. Canadian analysis of CSM in poultry diets has shown that, when supplemented with lysine, CSM reaches productivity levels comparable to SBM (Campbell, 1988).

4 MODEL DESCRIPTION Our study is grounded in the use of data obtained by feeding experiments 1 that employs full substitution of protein sources; i.e., the experiment is conducted with diets that contain either SBM or CSM for the collection of information on live body weight, feed consumption and weight of processed parts. This data set, which contains productivity information on each source (SBM vs. CSM), is utilized to estimate the production functions that will be employed in the profit maximization model of this study. A brief description of the model follows the mathematical specification: (1) subject to: (2), (3), (4), (5), (6). 1 Feed composition and feeding level experiment was conducted by Sterling et al. (2002). The experiments consisted of using four different levels of protein (17%, 20%, 23%, and 26%) and two different sources of protein (SBM vs. CSM) to feed broiler chickens until 42 days and collecting body weight, feed consumed and weight of whole carcass and cut-up parts.

5 In the objective function (Equation 1), maximum profit per bird per day (? ) for the integrator is defined as a function of derived price at farm gate 2 (DP BW ), live body weight (BW), cost of feed consumed (P FC ), feed consumed (FC), interest cost (I), and feeding time (t). The constraint set includes a number of physical factor constraint equations that, due to space limitations, are not explicitly discussed here. 3 However, the most relevant constraints that permit a direct comparison between the two sources of feed protein are described next. Cost of feed consumed includes feed delivery cost (DEL) and the least-cost feed (P F, Equation 2). The least-cost feed function (P F ) in Equation 3 minimizes the cost of feed for pre-determined ingredients (X i ) and their prices (P i ) and is determined by the optimization process underlying the least-cost feed formulation. Live chicken body weight (BW) is determined by feed consumed (FC), the square of feed consumed (FC 2 ), protein level (PR), square of protein level (PR 2 ), and an intercept shifter for female chickens (FE, Equation 4). The coefficients of the BW function are estimated by ordinary least squares on experimental data, and their values depend on whether SBM or CSM is chosen as the protein source. Further, coefficients in equations 5 and 6 are also estimated separately for SBM and for CSM using the experimental data. Feed consumed (FC) is determined by feeding time (t), feeding time squared (t 2 ), protein level (PR), protein level squared (PR 2 ), and an intercept shifter for female chickens (FE, Equation 5). Feed consumption is subject to a number of dietary, environmental, and genetic inputs. Experiment size constraints necessitated holding all factors constant except dietary protein, 2 See Costa et al. (2001) for details. 3 See Costa et al. (2002) for details.

6 amino acids, and gender. Results concur with those of Pesti and Smith (1984) that show that production responses of broilers to dietary energy and protein levels show diminishing marginal returns. Equation 6 is estimated as processed weight, W l, of each part l derived from a live bird (l = WC for whole carcass, BR for skinless boneless breast weight, TE for tenderloin, LQ for leg quarters, WI for wings, FP for fat pad, and RC for rest of chicken). The sum of all processed parts (plus offal and giblets) must equal the live weight of the bird. Each equation is estimated as a function of live bird weight (BW), protein level (PR), protein level squared (PR 2 ), and gender of birds (FE). The coefficients are estimated by OLS on experimental data, and their values are also posited to depend on whether SBM or CSM is chosen as protein source. Prices of inputs and outputs are collected for the profit maximization analysis. The price data consist of prices of ingredients available for the ration formulation, including major feedstuffs and synthetic amino acids that supplement the deficiencies of major alternative protein sources such as CSM. Prices received in Georgia (or the Southeast) for the outputs are considered in the analysis, as well as other costs considered in broiler production and processing 4. MODEL INTERACTIONS AND PRICE SENSITIVITY ANALYSES With the model developed above, we estimate the profitability of four base scenarios: broilers are produced and sold using either SBM or CSM as the protein source, and for each protein source, broilers are sold either after being processed into whole carcasses or into cut-up parts. Initially, 4 See Costa et al. (2002) for details.

7 comparisons are made directly between SBM vs. CSM results for each broiler product marketing alternative. Then, prices of inputs (SBM and CSM) are varied for input price sensitivity analysis. Lastly, a price mapping is extracted to indicate the impact on decisions of seasonal and other price variations that justify using CSM or SBM as the more profitable protein source. The solutions report broiler weight, feed consumption, feeding time, and feed composition that maximize profit under certain production function estimation, market option, and input/output prices. All optimal formulated rations meet or exceed minimum nutrient requirements from the National Research Council (NRC, 1994) for nutrient requirements in poultry production. Each optimized ration is fed to broilers for an optimized number of days in order to obtain a target weight that is to be processed and sold to a specific market, given the prices of outputs (processed parts) and inputs (feed ingredients, mainly), and other costs integrated in the model. Selling Broilers in the Whole Carcass Market The first analysis compares selling CSM- and SBM-fed broilers that are processed into whole carcass, with results presented in Tables 1 and 2 (first two columns). In Table 1, the feed formulated for CSM has less corn and more poultry fat than the SBM ration, but more CSM is used than for SBM rations in its composition (as CSM has a lower protein concentration than SBM). Note that both feed scenarios demonstrate optimal protein levels that are higher than the levels currently fed in the industry. CSM feed costs more per pound than the SBM feed in this product scenario (whole carcass), but the profit generated by supplying CSM to the broilers is higher than the profit generated by SBMfed broilers (Table 2). This represents the only feeding scenario among those tested at observed prices where CSM is more profitable than SBM. SBM-fed birds are fed longer (number of days) than are

8 CSM-fed birds and result in larger birds with higher consumed feed. Yet, profitability is higher for CSM-fed birds when measured as profit per broiler per day, profit per house per period, and profit per house per year. Table 1. Baseline Scenarios Used to Analyze the Profitability of CSM and SBM in Broiler Production, Whole Carcass and Cut-up Parts. Whole Carcass Cut-up Parts Variable Unit CSM SBM CSM SBM Protein Level % 23.92 23.12 24.23 23.92 Feeding Time days 34.94 39.80 34.94 40.07 Bird Weight lb 4.24 5.04 4.24 5.10 Feed Cost cents/lb 7.51 7.50 7.55 7.61 Feed Consumed lb/bird 6.32 7.97 6.30 8.04 Feed Conversion Ratio lb/lb 1.58 1.58 1.49 1.58 Profit (? ) cents/bird/day 2.08 1.52 2.30 2.59 Derived Price cents/lb 29.19 29.02 31.00 33.29 Broiler House Revenue $/house/period 12,778 9,234 14,710 19,622 Annual Profit $/house/year 116,600 74,898 134,227 159,159 Carcass Weight lb 2.90 3.55 --- --- Skinless Boneless lb --- --- 0.597 0.790 Breast Weight Tenderloin Weight lb --- --- 0.131 0.178 Leg Quarters Weight lb --- --- 1.335 1.625 Wings Weight lb --- --- 0.340 0.420

9 Table 2. Sensitivity Analysis of CSM Prices on Profitability of Broiler Production. Whole Carcass Variable Unit CSM Percentage Increase % from base 10 20 30 40 50 60 Protein Level % 23.65 23.36 23.24 23.02 22.81 22.61 Feeding Time (t) days 34.93 34.92 34.13 33.87 33.63 33.41 Bird Weight (BW) lb 4.23 4.23 4.09 4.04 3.99 3.95 Feed Cost (P F ) cents/lb 7.67 7.81 7.97 8.11 8.24 8.36 Feed Consumed (FC) lb/bird 6.33 6.34 6.04 5.94 5.86 5.79 Feed Conversion Ratio lb/lb 1.49 1.50 1.48 1.47 1.47 1.46 Profit (? ) cents/bird/day 2.05 2.02 1.98 1.95 1.92 1.89 Derived Price cents/lb 29.18 29.18 29.12 29.09 29.06 29.04 Broiler House Revenue $/house/period 12,472 12,169 11,553 11,168 10,803 10,459 Annual Profit $/house/year 113,804 111,041 108,128 104,517 101,102 100,457 Cottonseed Meal Price $/ton 159.50 174.00 188.50 203.00 217.50 232.00 It is important to note that the model shows the combination of many factors related to the production process: feeding as a function of time; body weight as a function of feeding; and, carcass weight as a function of live weight. These are the main factors that influence the production result. However, we must not forget that in considering CSM vs. SBM, relative cost is also a major factor. Selling Broilers in the Cut-up Parts Market Selling broilers processed into cut-up parts generates the highest profits overall (Table 2). Although this study s procedures differ significantly from the study conducted by Costa et al. (2001), the results obtained in this study are similar to theirs regarding the conclusion that SBM-fed broilers to

10 be sold as cut-up parts generally results in the most profitable scenario. The protein levels in the optimal solutions are again high relative to current industry practice, and the use of SBM and CSM in the diets for broilers processed as cut-up parts is at its highest level of all diets formulated for the marketing options simulated here. The feeding process is also the longest for this alternative, and feed costs are the highest of all SBM and CSM scenarios for processing alternatives. SBM-fed birds are fed longer than CSM-fed birds and result in larger birds with higher consumed feed. In this case, profitability is higher for SBM-fed birds when measured as profit per broiler per day, profit per house per period, and profit per house per year. Sensitivity Analysis of SBM and CSM Prices Economic feed formulations under CSM and SBM price variations are next analyzed for the carcass and cut-up parts marketing options. Because CSM-fed birds are more profitable than SBMfed birds for the whole carcass marketing option, the price of CSM is increased by 60% of the original price in increments of 10%, and profits for both protein sources are compared. The solutions generated over increased CSM prices are reported in Table 3. Results show that profit per bird per day decreases as CSM prices increase, but profit levels are still considerably higher than the profit level for SBM-fed birds reported in Table 2. The higher profitability of SBM-fed birds in the baseline cut-up parts processed birds scenario leads to a second comparison of the two protein sources when prices of SBM increase. SBM prices are increased by 70% over the original price in increments of 10%, and profits are compared to those of CSM-fed birds. Again, all variables reported in Table 3 decrease as SBM prices increase with the exception of feed cost. In this sensitivity analysis, however, notice that SBM profitability decreases

11 considerably and to levels approximating those from CSM-fed broilers. CSM-generated annual profits are lower than SBM-generated annual profits by less than 3,000 dollars in the final scenario (Tables 3 and 1). These results imply that CSM can be readily substituted for SBM in the cut-up parts processed broilers marketing option, but SBM is very unlikely to be substituted for CSM in whole carcass processed broilers marketing for relatively small percentage changes in the price of CSM. Table 3. Sensitivity Analysis of SBM Prices on Profitability of Broiler Production. Cut-up Parts Variable Unit SBM Percentage % from base 10 20 30 40 50 60 70 Increase Protein Level % 23.72 23.52 23.33 23.14 22.95 22.76 22.57 Feeding Time (t) days 39.85 39.65 39.46 39.28 39.12 38.97 38.84 Bird Weight (BW) lb 5.06 5.02 4.99 4.95 4.92 4.89 4.86 Feed Cost (P F ) cents/lb 7.80 7.98 8.15 8.31 8.47 8.61 8.75 Feed Consumed lb/bird 7.96 7.88 7.81 7.55 7.69 7.63 7.59 (FC) FCR lb/lb 1.57 1.57 1.57 1.56 1.56 1.56 1.56 Profit (? ) cents/bird/day 2.55 2.51 2.47 2.43 2.39 2.35 2.31 Derived Price cents/lb 33.24 33.20 33.15 33.10 33.06 33.01 32.96 Broiler House $/house/period 19,122 18,636 18,172 17,721 17,287 16.869 16,466 Revenue Annual Profit $/house/year 155,098 151,157 147,392 147,007 143,405 139,933 136,590 Soybean Meal Price $/ton 204.60 223.20 241.80 260.40 279.00 297.60 316.20

12 PRICE MAPPING ANALYSIS ON MEAL REGIMES At baseline prices for SBM and CSM, optimal solutions in Table 1 show that SBM generates higher profits than CSM when broilers are processed into cut-up parts, while CSM generates higher profits than SBM when broilers are processed into whole carcass. The prices of both sources can and do vary through time and/or seasonally, however. The analysis conducted here determines an equally profitable line for the price combinations of SBM and CSM. The interaction of many possible prices in the model determines at what price combination profits are equalized. Figures 1 and 2 display the isoprofit lines for the carcass and cut-up parts markets, respectively. Notice that in Figure 1, CSM-generated profits will be higher for price situations that fall below the iso-profit line, whereas SBM-generated profits will be higher for price situations that fall above the line. In the isoprofit line in Figure 1, equal profits are presented in points along the curve, and relatively greater profits for one or the other protein source result at greater distances from the curve. In Figure 2, notice that the line approximates a straight line, and for considerably higher prices of SBM than CSM, there will be equal profits. Again, CSM-generated profits will be higher for price situations that fall below the line, whereas SBM-generated profits will be higher for price situations that fall above the line.

13 Price Mapping for Isoprofit Levels of SBM vs. CSM (Carcass Market) CSM Price ($/ton) 600 500 400 300 200 100 0 0 50 100 150 200 250 Isoprofit SBM Price ($/ton) Figure 1. Isoprofit Price Mapping Analysis for Whole Carcass Marketing Option. Price Mapping for Isoprofit Levels of SBM vs. CSM (Cut-up Parts Market) CSM Price ($/ton) 350 300 250 200 150 100 50 0 0 100 200 300 400 500 600 SBM Price ($/ton) Isoprofit Figure 2. Isoprofit Price Mapping Analysis for Cut-up Parts Marketing Option.

14 CONCLUSIONS The profit-maximization model developed in this study interactively generates optimal solutions for CSM- and SBM-fed broilers in the whole carcass and cut-up parts marketing options. Feeds formulated for optimal solutions meet or exceed minimal NRC (1994) requirements for nutrient composition of feed rations. Protein levels of the optimal feed rations indicated by the model are above the average levels reported in the industry and range from 23% to just less than 25% of the diet. Profits are higher for CSM-fed broilers when marketed in the whole carcass marketing option, whereas profits are greater for SBM-fed broilers when marketed in the cut-up parts option. However, CSM-fed birds are fed for shorter feeding times than are SBM-fed broilers in all directly compared scenarios, with average live body weight and total feed consumed also being lower for CSM-fed broilers than for SBM-fed broilers. A price mapping analysis of SBM vs. CSM profits for the whole carcass and cut-up parts markets indicates that there are price combinations at which both sources are equally profitable. The range of prices for which CSM is more profitable than SBM is relatively larger in the whole carcass market, indicating that CSM can more-profitably be used for feeding broilers in that market. On the other hand, the range of prices for which SBM is more profitable than CSM is relatively larger in the cut-up parts market, indicating that SBM will generally be more profitable than CSM for feeding broilers in that market. Importantly, results for both SBM and CSM formulated rations also show that poultry producers could increase profitability by formulating rations that have higher protein levels than the currently recommended levels. Solution sets obtained from profit maximization model interactions demonstrate that SBM-

15 based diets are generally more efficient than CSM-based formulations. Moreover, at the set of input and output prices extant, SBM-based feeds are more profitable than using CSM, especially for selling broilers processed into cut-up parts. CSM can be fed more profitably than SBM at these input and output prices only when broilers are sold as whole carcasses. CSM may thus have a potentially new market the protein input market for poultry production in the Southeast and other areas of potential excess CSM. That is, given its availability, the results provided in this study demonstrate that CSM may be used profitably as an alternative protein source for SBM in broiler production. REFERENCES Anderson, J. O., and R. E. Warnick. Sequence in which Essential Amino Acids Become Limiting for Growth of Chicks Fed Rations Containing Cottonseed Meal. Poultry Science 45(1966):84-89. Campbell, L. D. Canola Meal as a Substitute for Cottonseed Meal in the Diet of Broiler Chickens. Nutrition Reports International 37(1988):371-377. Costa, E. F., B. R. Miller, J. E. Houston, and G. M. Pesti. Production and Profitability Responses to Alternative Protein Sources in Broiler Rations. Journal of Agricultural and Applied Economics. 33.3(2001):567-581. Costa, E. F., J. E. Houston, L. Gunter, and G. M. Pesti. "Incorporating Consumer Demand in Broiler Profit-Maximization Modeling." Journal of Food Distribution Research 33(2002):1-10. Feedstuffs: The Weekly Newspaper for Agribusiness. Ingredient Market. v. 72, n. 47, November 13, 2000.

16 Fisher, H., and J. H, Quisenberry. Net Protein Utilization and Amino Acid Deficiencies of Glandless Cottonseed Meal. Poultry Science 50(1974):1197-1200. Grau, C. R. Protein Concentrates as Amino Acid Sources for the Chick: Corn Gluten Meal, Cottonseed Meal and Peanut Meal. Journal of Nutrition 32(1946):303-312. National Cottonseed Products Association, NCPA. Cottonseed Products Yield Table, 1999. http://www.cottonseed.com. National Research Council (NRC). Nutrient Requirements of Poultry. (9th revised edition). Washington, DC: National Academy Press,1994. Pesti, G. M., and C. F. Smith. The Response of Growing Broiler Chickens to Dietary Protein, Energy and Added Fat Contents. British Poultry Science 25(1984): 127-138. Sterling, K. G., E. F. Costa, M. H. Henry, G. M. Pesti, and R. I. Bakalli.. "Responses of Broiler Chickens to Cottonseed- and Soybean-Meal Based Diets at Several Protein Levels." Poultry Science Journal 81(2002):217-226.