Download PDF

Hog Producer Investment in
Value-Added Agribusiness: Risk and Return Implications
Brian R. Jones
Supply Planning
Pioneer Hi-Bred Int'l. Inc.
Joan R. Fulton
Department of Agricultural Economics
Purdue University
Frank J. Dooley
Department of Agricultural Economics
Purdue University
Selected Paper at
American Agricultural Economics Association Annual Meeting,
August 8-11, 1999, Nashville, TN
Copyright 1999 by Brian R. Jones, Joan R. Fulton and Frank J. Dooley. 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.
Hog Producer Investment in
Value-Added Agribusiness: Risk and Return Implications
Abstract
Although producers have been enticed into investment in value-added agribusiness the
risk and return impacts have not been quantified. A spreadsheet simulation model is used
to evaluate how investments by hog farmers in slaughter plants and other alternatives
affect returns and risk. Results suggest that hog producer investment in value-added
agribusiness is efficient.
There has recently been considerable investment in value-added agribusiness by
producers in all sectors of agriculture. Producers are enticed and sometimes lured to
move along the value chain with promises of higher profit margins and lower levels of
risk through investments in value-added processing (Harris, Stefanson, and Fulton).
Investments in value-added agribusiness have taken many different forms, including
producer networks, limited liability companies, limited partnerships, and New Generation
Cooperatives.
Hog producers have been particularly interested in value-added investment
opportunities in hog slaughtering for two reasons. First, industrialization in the hog
industry is occurring at an increasing pace (Boehlje, et. al., Rhodes). Thus, producers
feel they must restructure their business to survive. Second, the recent crisis in the pork
industry, with hog prices in December 1998 at levels lower than prices of the Great
Depression, is in part being blamed on inadequate capacity in hog slaughter facilities.
While the opportunity to take advantage of higher returns and lower risk through
investments in value-added processing is extremely appealing, it demands further
analysis. It is not obvious that farmers' investment in value-added processing of their
commodities is the most advantageous from the perspective of their financial portfolio.
Since the same underlying demand factors may affect the raw commodity and the
processed value-added product, it may be that the income stream of the farm operation is
positively correlated with the income stream of the value-added processing firm. If this
is the case, farmers could end up increasing their exposure to risk by investing in valueadded processing and could more effectively diversify their portfolios by making
alternative investments. Conversely, value-added investments may be warranted if the
1
income streams are independent or negatively correlated (Fulton and Boehlje). This
research examines the financial risk and return implications of investment in value-added
processing for hog producers.
The following section of the paper describes the simulation model that was
developed to evaluate the risk/return tradeoffs. The model incorporates the hog farm
operation, a pork slaughter plant, and investment possibilities in stocks and bonds.
Section three of the paper describes the alternative investment scenarios that a hog
producer might consider. The fourth section of the paper presents the results of the
simulation analysis and compares the alternatives using the following efficiency criteria:
(i) Expected Value/Variance, (ii) First Degree Stochastic Dominance, and (iii) Second
Degree Stochastic Dominance. The final section of the paper contains conclusions and
suggestions for further study.
Pork Production Investment Portfolio Model
A simulation model incorporating the hog farm operation, a pork slaughter plant,
and investment possibilities in stocks and bonds was developed to examine the risk/return
tradeoffs facing hog producers. The model was developed in Microsoft Excel, using
@Risk to incorporate the stochastic nature of the investments. The model has four
components that calculate return on investment for the hog farm operation, the pork
slaughter plant, investments in stocks, and investments in bonds.
The hog farm portion of the model is based on the assumption that the producer
operates a farrow-to-finish operation. Farm Profit is calculated for each of three different
sizes of hog operations (300-sow, 600-sow and 1200-sow) according to the following
equation:
2
1. Farm Profit = Farm Revenue - Feed Costs - Other Variable Costs - Fixed Costs
Farm Revenue is calculated in the standard way by multiplying the number of
hogs produced by the average live weight of slaughtered hogs by the price of finished
hogs. Corn and soybean meal constitutes the Feed Costs in this model and make up the
largest portion of variable costs. The Other Variable Costs include feed supplements,
health supplies, fuel, utilities, marketing, mortality disposal, and artificial insemination
costs. This value is calculated by multiplying the number of pounds of live hogs
produced by a variable cost per hundredweight value. Fixed costs in this model include
land, building, equipment, labor and management, as well as the sows used in production
and a portion of the market inventory that is on hand at any given time. The cost of
production estimates and associated economics and production assumptions are taken
from Foster, Hurt, and Hale (1995).
Farm return on investment (ROI) is calculated by dividing farm profit by the total
capitalized value of the farm. The total capitalization for the farm represents a snapshot
of the operation running at full capacity on any given day and is a sum of the land,
buildings, equipment, production inventory, and marketing inventory. These
capitalization estimates were reported by Foster, Hurt, and Hale (1995).
The pork slaughter plant component of the model simulates revenues and fixed
and variable costs for a plant with capacity of two million head per year, the current
industry standard. The cost and revenue information associated with processing hogs is
not as readily available as the cost and revenue information for farm production. Two
sources of cost information were used to develop this section of the model. Hayenga's
research involving interviews with managers from eight hog processing firms provided
3
one source. Dooley's disaggregation of the Census of Manufactures data to determine
variable costs was the second source. The profit from the processing plant is determined
by the following equation:
2. Processing Profit = Processing Revenue - Livestock Cost - Variable Costs - Fixed Cost
Processing revenue is made up of three parts: cutout revenue, byproduct revenue,
and fat revenue. The pork cutout revenue is the most complicated to calculate. The
average live weight of slaughtered hogs is multiplied by the carcass yield factors to
determine the average cutout weight of a slaughtered hog. A value for total carcass cutout
pounds is then determined accounting for the number of hogs slaughtered. Finally cutout
revenue is calculated by multiplying total carcass cutout pounds by pork carcass cutout
value. Byproduct revenue is calculated by multiplying the byproduct credit by the
number of hogs slaughtered. Finally, fat revenue is calculated by multiplying the fact
credit by the number of hogs slaughtered.
Since the livestock costs are a major component of variable costs for a hog
processor they were estimated separately from other variable costs. This is calculated the
standard way by multiplying the total live pounds purchased by the price of finished
hogs. The term "Variable Costs" in the above equation represents the costs associated
with operating a slaughter plant other than livestock costs and fixed costs. This value is
calculated by multiplying the number of hogs slaughtered by a variable cost per head as
reported by Hayenga. Fixed costs include plant and equipment costs, interest on
investment, property tax, and insurance on the hog slaughter facility. Fixed costs are
estimated on a per head basis and multiplied by the maximum capacity of the processing
facility.
4
Processing return on investment (ROI) is calculated by dividing processing profit
into the total capitalized value of the processing facility. The capitalization for a
processor is the cost to construct a processing facility with an annual capacity of two
million head (Jones, 1998).
Two investment opportunities, beyond the farm and pork processing, made up the
third and fourth components of the model.. In the first case, it was assumed that the
producer could invest in the stock market. To simulate returns for stock market
investments historical returns on the S&P 500 were used. Second, the possibility of
investing in bonds was incorporated into the model using historical returns from Treasury
Bills. To make the model more realistic, appropriate probability distributions were
determined for key variables and incorporated into the model using the @Risk
software. Table 1 outlines the stochastic variables used in the model and the type of
distribution that best fit the data. In addition, the historical source and time series
corresponding to each input variable is included. Note that the data were not adjusted for
inflation, so the returns generated by the model are nominal rates of return. Since
historically, some of the variables are correlated this was incorporated into the @Risk
model. The variables that were correlated in the model are the price of hogs, corn, and
soybean meal, and pork carcass cutout value. The correlation values used in the model
are: price of hogs to price of corn of .0159, price of hogs to price of soybean meal of
.2100, price of hogs to pork carcass cutout value of .9561, price of corn to price of
soybean meal of .3803, price of corn to pork carcass cutout value of -.0544, and price of
soybean meal to pork carcass cutout value of .2131.
5
Scenarios
Six discrete investment portfolio alternatives available to hog producers were
compared in this analysis (Table 2). These portfolio alternatives are somewhat arbitrary,
but represent potential investments that hog producers might consider. The alternatives
were chosen with the assumption that farmers were not likely to allocate more than 25
percent of their portfolio to off-farm investments. The return on investment (ROI) for
each of the alternatives was calculated by taking a weighted average of the ROI for the
individual investments. In examining the six alternative scenarios, it was assumed that
both the farm operation and the pork processing plant are going concern businesses that
are already operating efficiently at the desired capacity. In addition, it was assumed that
producers could find investment opportunities for the exact amount of capital that they
had available for investment and did not have to worry about not having enough capital
for a particular project. Future study could examine additional considerations such as the
total amount of capital needed, cash flow requirements and net present value impacts of
building one or both of these operations new with cash outlay necessary in the early years
followed by revenue streams in the later years.
Results
The model was run using @Risk to perform Monte Carlo simulations. A
benefit of @Risk output is that it gives more detailed information than just a point
estimate. In this model, the output not only includes an expected (mean) return, but it
also provides a standard deviation of the return as well as each of the return values from
the number of iterations. The number of iterations, in this analysis, was determined by
the convergence criterion that all output percentage changes were less than 1.5 percent.
6
This resulted in 2200 iterations for the 300-sow operation, 1500 iterations for the 600sow operation, and 2200 iterations for the 1200-sow operation.
Since this model calculates the return for six discrete investment alternatives,
distributions and return values are available corresponding to each investment alternative.
These return values enable comparison between each investment based on the efficiency
criteria of: Expected Value/Variance, First Degree Stochastic Dominance and Second
Degree Stochastic Dominance.
A summary of the results, including the mean ROI, standard deviation, minimum
ROI, and maximum ROI, of the model runs are presented in Table 3. The six alternative
scenarios available to the producer are presented in the columns. The highest ROI for
any of the six scenarios is for the 1200-sow operation, demonstrating the size economies
that exist in hog production today. It is not surprising that the lowest return is for the
alternative with 75% investment in the farm and 25% investment in Treasury Bills for
any size of hog operation. The scenario with the highest ROI, for hog operations of 300or 600-sows, is 75% investment in the farm and 25% investment in processing.
However, for the 1200-sow operation the highest ROI is for 100% investment in the
farm. The scenario with the largest standard deviation, or greatest level of risk, is 100%
investment in the farm for all sizes of hog operations. The scenario with the lowest
standard deviation, or lowest level of risk, is 75% investment in the farm, 15%
investment in processing, and 10% investment in Treasury Bills.
The six alternative scenarios were compared by the three efficiency criteria as
reported in Table 4. The efficiency criteria are very discriminating for this model. The
Expected Value/Variance criteria was able to categorize five of the scenarios in the
7
inefficient set, leaving only the 75% investment in the farm, 25% investment in
processing in the efficient set for the 300- and 600-sow operations. In the case of the
1200-sow operation, the 100% investment in the farm scenario was also left in the
efficient set. However, many investors would select the 75% investment in the farm,
25% investment in processing over 100% investment in the farm since the drop in the
mean is relatively small (from 20.19 to 19.68) while the decrease in the standard
deviation is substantial (from 13.19 to 8.80).
The First Degree Stochastic Dominance efficiency criteria is not usually very
discriminating and this is no exception with only one scenario placed in the inefficient set
in the 600-sow operation and two scenarios in the inefficient set in the 300-sow and
1200-sow operations. However, the Second Degree Stochastic Dominance efficiency
criteria are discriminating. In the case of the 300-sow operation, the only scenario left in
the efficient set is 75% investment in the farm and 25% investment in processing. The
75% investment in the farm and 25% investment in processing, as well as the scenario
involving investment in the farm, processing and the S&P500 are left in the efficient set
for the 600-sow operation. For the 1200-sow operation, the 100% investment in the farm
along with the 75% investment in the farm and 25% investment in processing are left in
the efficient set. The fact that scenarios involving investment in processing consistently
remain in the efficient set, for different efficiency criteria and different sizes of hog
operations provides support for value-added investments by hog producers.
Conclusions and Suggestions for Further Study
In summary, diversification is a strategy hog producers should consider.
However, it is important to note that as the farmer’s size of operation increases, the
8
attractiveness of other investment alternatives such as hog processing, the S&P 500, and
Treasury Bills diminishes. This is a result of the gain from the increased efficiencies
achieved by the larger hog producers.
It is important to note that the liquidity preferences of the farmer investors were
not considered when comparing the investment alternatives in this research. Value-added
hog processing is less liquid than Treasury Bills or a stock portfolio. While investors can
easily enter and exit the stock or bond markets investment in a hog processing facility
usually requires a long-term commitment. In many of the cooperatively owned
processing facilities long-term delivery contracts and equity investments are required by
the owner investors. Therefore, if maintaining a high degree of liquidity is very
important for a producer diversification in hog processing may not be an attractive
alternative.
As noted earlier in the paper, this analysis assumed that the hog production and
hog processing operations are going concern businesses and that a producer is able to
invest any amount of money in the businesses. Future research could relax these
assumptions and examine the financial implications concerning equity required for
purchase of a hog processing facility, cash flow, liquidity, and leverage requirements.
Finally, the impact on market prices of more tightly aligned supply chains has not
been considered, and represents an opportunity for future study. As supply chains become
more tightly aligned with contracts and other negotiated arrangements, markets become
"thin." Little is known about the impact of these changes on price discovery of
agricultural commodities, and in turn, on the implications for producer investments in
value-added processing activities.
9
Table 1. Stochastic Variable Distributions and Historical Data Source
Variable
Distribution
Historical Data Source
Price of Finished Log Normal Truncated
USDA/AMS: Jan 1980 – June
Hogs ($/cwt)
(47.03, 7.04, 15.00, 63.44)
1998 monthly prices for barrows
(Mean, Std Dev, Min, Max)
and gilts in IA/MN
Price of Corn
Log Normal Truncated
USDA/AMS: Jan 1980 – June
($/bushel)
(2.57, 0.58, 1.34, 4.86)
1998 monthly prices for corn #2
(Mean, Std Dev, Min, Max)
yellow, Central Illinois
Price of Soybean Log Normal Truncated
USDA/AMS: Jan 1980 – June
Meal
(199.53, 40.35, 119.40,
1998 monthly prices soybean
($/ton)
306.39)
meal, 48% solvent, Decatur
(Mean, Std Dev, Min, Max)
Illinois
Average Live
Normal Truncated
USDA/NASS: Jan 1980 – June
Weight of
(249.18, 5.47, 239.00, 261.00) 1997 monthly average live weight
Slaughtered
(Mean, Std Dev, Min, Max)
of hogs slaughtered under federal
Hogs (pounds)
inspection
Pork Carcass
Normal Truncated
USDA/NASS: Jan 1980 – June
Cutout Value
(63.52, 8.10, 41.25, 84.32)
1998 monthly pork carcass cutout
($/cwt)
(Mean, Std Dev, Min, Max)
value, U.S. No. 2.
Carcass Yield
Normal
Calculated from USDA/ERS/AMS
(ratio)
(0.7301, 0.0055,)
data 1996-1997 monthly carcass
(Mean, Std Dev)
yields.
Processor
Triang
Hayenga: Survey of Pork
Variable Cost
(20,22,25)
Processing Plant Costs. Reported
($/head)
(Min, Most Likely, Max)
range of variable cost for single
shift plant.
Processor Fixed Triang
Hayenga: Survey of Pork
Cost
(3,6,10)
Processing Plant Costs Reported
($/Head)
(Min, Most Likely, Max)
range of fixed cost for single shift
plant.
Pork Processing Normal
U.S. Census Bureau: Survey of
Capacity
(1,685714, 34087)
Plant Capacity Utilization. 1989(million
(Mean, Std Dev )
1996
head/year)
S&P 500
Normal
Yahoo Finance: Jan 1980 – June
(%)
(12.80, 14.30)
1998 calculated annual return
(Mean, Std)
based on monthly closing prices
Treasury Bills
Log Normal Truncated
Federal Reserve Board of
(%)
(6.98, 2.98, 2.86, 16.30)
Governors: Jan 1980 - June 1998
(Mean, Std Dev, Min, Max)
Monthly Treasury Bill Rate
10
Table 2. Alternative Producer Investment Scenarios
Investment Scenario
Acronym
100% investment in the farm
FARM
75% investment in the farm and 25% investment in pork processing
FP7525
75% investment in the farm and 25% investment in S&P 500
FSP7525
75% investment in the farm and 25% investment in Treasury Bills
FTB7525
75% investment in the farm, 15% investment in pork processing, and 10%
investment in S&P 500
75% investment in the farm, 15% investment in pork processing, and 10%
investment in Treasury Bills
FPSP
Table 3. ROI Results of the Simulation Model
FARM FP7525 FSP7525
FPTB
FTB7525
FPSP
FPTB
300-sow
Mean
10.41
12.16
11.08
9.55
11.73
11.12
Farrow-
Std. Dev.
11.39
7.75
9.26
8.56
7.97
7.84
to- Finish
Minimum
-27.03
-12.13
-20.98
-18.33
-15.67
-14.03
Operation Maximum
43.25
36.21
38.89
35.10
36.97
35.38
600-sow
Mean
16.70
16.99
15.76
14.23
16.50
15.89
Farrow-
Std. Dev.
12.24
8.41
9.95
9.21
8.68
8.54
to- Finish
Minimum
-21.09
-12.07
-18.72
-13.46
-11.15
-12.63
Operation Maximum
52.77
41.90
48.19
41.30
42.14
41.39
1200-sow
Mean
20.19
19.68
18.36
16.87
19.15
18.55
Farrow-
Std. Dev.
13.19
8.80
10.60
9.90
9.17
9.04
to- Finish
Minimum
-22.52
-11.74
-16.03
-15.69
-13.45
-13.32
Operation Maximum
59.25
47.88
52.06
47.00
48.33
47.29
11
Table 4. Efficiency Criteria Summary
Expected Value
(Mean) Variance
2nd Degree Stochastic
Dominance
Efficient
Set
Inefficient
Set
Efficient
Set
Inefficient
Set
Efficient
Set
Inefficient
Set
FP7525
FARM
FSP7525
FTB7525
FPSP
FPTB
FARM
FP7525
FSP7525
FPSP
FTB7525
FPTB
FP7525
FARM
FSP7525
FTB7525
FPSP
FPTB
FP7525
FARM
FSP7525
FTB7525
FPSP
FPTB
FARM
FP7525
FSP7525
FPSP
FPTB
FTB7525
FP7525
FPSP
FARM
FSP7525
FTB7525
FPTB
FARM
FP7525
FSP7525
FTB7525
FPSP
FPTB
FARM
FP7525
FSP7525
FPSP
FTB7525
FPTB
FARM
FP7525
FSP7525
FTB7525
FPSP
FPTB
300-sow
Farrow-toFinish
600-sow
Farrow-toFinish
1200-sow
Farrow-toFinish
1st Degree Stochastic
Dominance
12
References:
Boehlje, Mike, Kirk Clark, Chris Hurt, Don Jones, Alan Miller, Brian Richert, Wayne
Singleton, and Allan Achinckel. (1997). Hog/Pork Sector. Food System 21:
Gearing up for the New Millennium. EC 710, Purdue University Cooperative
Extension Service: West Lafayette, Indiana.
Boehlje, Michael, Jay Akridge, and Dave Downey. "Restructuring Agribusiness for the
21st Century." Agribusiness.11(6):493-500, 1995.
Dooley, Frank. J. (1999) "Estimating Plant Level Costs For Food Processors With Census
Of Manufacturers Data." Staff Paper. Dept. of Agricultural Economics, Purdue
University, West Lafayette, IN.
Foster, Kenneth, Chris Hurt, and Jeffrey Hale. (1995) Positioning Your Pork Operation
for the 21st Century. Purdue University Cooperative Extension Service: West
Lafayette, IN.
Fulton, Joan R. and Michael Boehlje. (1998) The Cooperative Choice: Strategy or
Anachronism." Paper presented in Organized Symposium at 1998 annual meeting
of American Agricultural Economics Association, Salt Lake City, Utah.
Harris, Andrea, Brenda Stefanson, and Murray Fulton. (1996). New Generation
Cooperatives and Cooperative Theory. Journal of Cooperatives, 11, pp. 15-28.
Hayenga, Marvin L. (1997) Cost Structures of Pork Slaughter Processing Firms:
Behavioral and performance Implications. Iowa State University Staff Paper #
287.
Jones, Brian R. (1999) The Risk and Return Implications of Value-Added Investment by
Hog Producers. Masters Thesis, Purdue University.
Jones, Rick (1998). www.ads.uga.edu/groups/swine/coop1.htm . Sunbelt Pork
Cooperative. (Accessed on 10/23/98).
Rhodes, V. James. (1995). The Industrialization of Hog Production. Review of
Agricultural Economics, 17, 107-118.
13