76
CREIGHTON UNIVERSITY FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965 HARRY T. AMBROSE

FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

  • Upload
    others

  • View
    5

  • Download
    0

Embed Size (px)

Citation preview

Page 1: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

CREIGHTON UNIVERSITY

FORECASTING MILLFEED PRICES

MBA 299

SEPTEMBER 7, 1965

HARRY T. AMBROSE

Page 2: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

ALUiriNi MEMORIAL LIBRARY Creighton University /

| Omaha, Nebraska 68131 303684

Page 3: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

CONFIDENTIAL

Please treat all the information on Nixon feed tonnages and percentages as confidential. Please handle the data on feed tonnage of members of the American Feed Manufacturers Association in the same manner.

11

Page 4: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

ACKNOWLEDGMENTS

The author is indebted to many people for help­ing to produce this paper. Chief among them is my wife, Margie, who made graphs, kept the "office" quiet, gave up bridge games and buoyed my morale.

Dottie Sorensen spent many hours typing this, and I am indeed indebted to Nebraska Consolidated Mills for allowing me to use its resources in this effort.

Tom James, Wally Bierman and Marv Wilkening of the NCM nutrition department helped me with advice and information.

Mr. Oakley Ray of the American Feed Manufacturers Association kindly furnished AFMA sales tonnage analyses. John Branston of General Mills provided information on industrial usage of millfeeds. Neil Sosland of The South­western Miller dug up the urea article for me.

Walter T. Borg and H. William Overbeck, Jr., of the U.S.D.A. provided specific information. William Briscoe of the Kansas Crop and Livestock Reporting Service

iii

Page 5: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

spent a good deal of time with me on the telephone fill­ing in gaps in my animal number statistics. Professor Gleason of Creighton provided overall guidance on the project.

iv

Page 6: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

CONTENTS

PageCONFIDENTIAL .............................. iiACKNOWLEDGEMENTS .......................... iiiLIST OF TABLES............................ vLIST OF GRAPHS ............................... viiChapter

I. HYPOTHESIS........................ 1II. PRODUCTION........................ 8III. NUTRITIVE ANALYSIS ................ 11

General Layer Feeding Turkey Feeding Hog Feeding Cattle Feeding Dairy Feeding Conclusion

IV. COMPETING INGREDIENTS ............ 37V. WEIGHTING ANIMAL NUMBERS .......... 40VI. THE CORRELATION ANALYSIS........... 55

BIBLIOGRAPHY 66

Page 7: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

515161721

21

2525

272833

3335414546

LIST OF TABLES

U. S. Millfeed Tonnage by Quarters . .Nutrient Analysis ..................Average Prices by Quarter ..........

0

Ingredient Table ....................Nixon Poultry Feeds ................Percentage for Layer Type Feeds at Nixon, Omaha ......................

"Typical" Percentage Sold Data . . . ."Typical" Turkey Feeds ..............Percentage of Each Type Feed Sold at Nixon, Omaha ......................

Nixon Swine Feeds ..................Nixon Beef Feeds ....................Percentage of Type Feed Sold at Nixon, Omaha ............................

Nixon Dairy Feeds ..................A.F.M.A. Feed Tonnage ..............Inventories of Hogs on Farms ........Turkeys Produced ....................

v

Page 8: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

Table Page

17. Inventories of Other Cattle on Farms . 4718. Inventories of Milk Cows on Farms . . 4819. Inventories of Layers on Hand . . . . 4920. Feed Tonnage, Quarter Weighting . . . 5121. Weighted Animal and Poultry Numbers . 5322. The Correlation Analysis............ 56

Page 9: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

LIST OF GRAPHS

Graph Page1. Layer Feed Tonnage.................... 222. Nixon Urea Based Cattle Tonnage . . . 323. Bulk Prices FOB Kansas City, Mo. . . . 384. Weighted Average Price Soy, CSM, Milo,

C o r n .................................395. K.C. Millfeed Average Prices Versus

Millfeed Domestic Disappearance . . 586. K.C. Millfeed Average Prices Versus

Weighted Competing IngredientPrices.............................. 59

7. K.C. Millfeed Price Versus WeightedAnimal and Poultry Numbers ........ 60

8. Basis for Computing Column 8 ........ -619. Basis for Computing Column 10 . . . . 62

vii

Page 10: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

CHAPTER I

HYPOTHESIS

Nebraska Consolidated Mills is my company. Our business is food. Nebraska Consolidated Mills came into existence as wheat millers, and over the years have also become manufacturers of animal feed and growers of chickens, hogs, and cattle.

All of these activities have common threads link­ing them into logical extensions of each other. One ofthese is wheat millfeeds, sometimes termed a by-product

1and, more recently, a co-product of wheat milling. Its principal usage is as a feed ingredient, and it is a very important one.

Millfeeds take several forms, red dog, bran, shorts, middlings, etc. They constitute 25-30% of the total amount of wheat processed on the basis of weight.

^"Millfeed, By-Product, Co-Product or Product?" Reprint of panel discussion presented at the 61st Annual Convention of the Millers' National Federation, May 14, 1963. Available from the M.N.F. office, 309 W. Jackson Blvd., Chicago, 111., p. 2.

-1-

Page 11: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-2-

The percentage depends principally on the type of wheat and efficiency of the mill. Their value varies consider­ably but will average 15-20% of the total return from the wheat milled.

Feed millers utilize all of these products to a greater or lesser degree in manufacturing poultry and animal feeds of all kinds. Historically, millfeeds were probably the first commodity other than grains and forage crops in their natural state to be fed to livestock.

The purpose of this paper is to explore selected aspects of millfeed marketing to as great a depth as possible with the specific end in view of developing multiple regression and correlation relationships which might be helpful in forecasting the price of middlings at Kansas City.

This market has by far the greatest relevancy for both the Feed and Flour Divisions of our company.All of our flour mills make one millfeed product, standard wheat middlings. They are all located in areas where the Kansas City market has the greatest effect on price of any major market.

The prime Kansas City market area is Kansas, Nebraska, Iowa, and Missouri. Basically speaking, the

Page 12: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-3-

flour mills and feed mills in this area are in close balance relative to supply and demand for millfeed.With minor exceptions almost all millfeed produced hereis consumed here and vice versa. The minor exceptions

2probably exert a great effect on price, however.Nebraska Consolidated Mills feed mills are also

located within the sphere of the Kansas City market.*

Nixon and Co., a wholly owned subsidiary, operates in Iowa and Nebraska with 5 mills, the largest of which is located in Omaha. In developing demand factors for mill- feeds, Nixon's Omaha experience will be extensively used.

The forecast will be developed by examining factors of supply and demand, stating an hypothesis, and then showing the relationship between the sets of data to see if it can be proven. In addition, explana­tory material will be presented to further the understand­ing of the reader into the underlying factors in the overall situation. Liberal use will be made of tables and graphs in this area.

"Millfeed Production and Marketing Practices in Kansas and Missouri," Agri-Research, Inc., Manhattan, Kansas, March, 1965, p. iv.

Page 13: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-4-

Flour mills have very little capability to store millfeeds. That which does exist is usually limited to warehouses for sacks. The trend to bulk handling continuously causes this to shrink. Nebraska Consolidated Mills flour mills will average less than 20 tons sacked storage at all plants at any given time which is about 2 hour'sproduction. Nebraska Consolidated Mills and most of its

*competitors simply blow bulk millfeeds into railroad cars from the production stream. The cars are shipped directly to buyers. This may be verified by the infor­mation on month end stocks in Table 1.

Millfeed production is very constant over the seasons, but since feed demand is not constant, millfeed must have relatively large price fluctuations in order to ration its use.

The animal and poultry categories which are major feed consumers in the area included are beef cattle, swine, dairy cattle, laying chickens and turkeys. Each of these has a definite seasonal variation in feed tonnage requirements. The reader is invited to refer to Table 5 for an analysis of this. The different animal

Ibid., p . 8.1

Page 14: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-5-

TABLE 1U.S. MILLFEED TONNAGE BY QUARTERS

000‘S OMITTED

Date Production Imports ExportsDomestic

Disappearance

J-M 65 1,060 13 6 1,0680-D 64 1,274 20 11 1,283J-S 1,073 25 7 1,091A-J 1,327 42 4 1,365J-M 1,148 73 3 1,2180-D 63 1,263 67 4 1,326J-S 1,223 18 8 1,233A-J 1,189 17 13 1,193J-M 1,203 24 10 1,2170-D 62 1,244 20 12 1,252J-S 1,207 16 12 1,211A-J 1,150 12 16 1,146J-M 1,275 17 9 1, 2830-D 61 1,274 21 3 1, 292J-S 1,239 11 3 1,247A-J 1,112 12 9 1,115J-M 1, 233 16 5 1,2440-D 60 1,294 13 9 1,298J-S 1,205 15 13 1,207

Notes:There was no carryover for any of the quarters

listed.aTaken from: "Feed Market News, Weekly Summary

and Statistics," various issues 1961-65, Grain Division, A.M.S., U.S.D.A., Washington, D.C.

Page 15: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-6-

types utilize millfeeds in varying percentages in their feeds. One of the major determinants of the percentage is price of millfeeds and competing ingredients.

It is possible to develop equations through the use of multiple regression and correlation techniques which will have a high coefficient of correlation in explaining the movement of the dependent variable, mill- feed price. There are probably three independent variables which must be used. The first is supply of millfeeds, the second and third are representative of demand. One of these is the number of animal and poultry consuming units in existence and the other is the price structure for competing ingredients.

In demonstrating this relationship, extensive data will be used and explanatory material presented.The explanatory material is designed to give the reader an insight into some of the conditions present in the area of millfeed marketing.

Chapter II develops the production of millfeeds and its effect on the market situation. Similarly,Chapter III examines the animal and poultry feeding situation as a guide to the demand factor. Chapter IV introduces the element of feed ingredients which compete

Page 16: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-7-

with millfeeds for a place in the formulations of finished feeds. Chapters V and VI actually develop the forecast by weighting the data and discovering the relationships which exist.

Page 17: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

CHAPTER II

PRODUCTION

Each month the United States Department of Agriculture publishes in Feed Market News statistics detailing total millfeed production, imports, exports, and usage for the entire United States. These statistics are gathered from Bureau of the Census data, which is considered to be very accurate. The figures are indi­cated in Table 1.

For several reasons millfeeds are consumed shortlyafter they are produced so there have never been anystocks carried forward from one month to th-e next. Thisis borne out by the study done by Agri-Research, Inc.,in "Millfeed Production and Marketing Practices in Kansas

1and Missouri."

As indicated in the hypothesis, millfeed produc­tion is very steady as is domestic disappearance. Imports and exports vary quite a bit, however, and probably exert a disproportionate effect on the market.

1Ibid.

Page 18: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-9-

Imports, in particular, almost universally come from Canada and enter the northeastern states of Vermont, New Hampshire, Maine, Massachusetts, and Connecticut.None of those states has a flour mill, and they tradition­ally draw U.S. production from mills in Buffalo, New York. Heavy imports put pressure on Buffalo prices which tend to decrease the premium for Buffalo millfeed over the markets further west.

Recently, a large sale of flour by Canada to Russia was announced. Buffalo millfeeds prices immedi­ately declined $5 per ton. Kansas City was unchanged.

Exports have been small until this year. They have gone mainly from Seattle, Washington and Portland, Oregon to Japan. This year over 25,000 tons have been sold to countries in the European Common Market. This is because the European Common Market applies an ad val­orem duty of $20-$30 per ton to any feed ingredients whose starch content exceeds 28 per cent. For millfeedswith less than that percentage, the duty is $5-$8 per

2ton. This situation gives millfeeds an advantage over

Herman Steen, "A Promising Outlet for Millfeed," The Southwestern Miller, July 13, 1965, page 32.

Page 19: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-10-

cereals such as milo and corn and proteins such as soy­bean meal. If it continues, the United States should export considerable tonnage to the European Common Market in the future. In order to have the high density required for economical ocean shipping, export millfeeds are always pelletted. x

Production of millfeeds can be forecast with accuracy up to a year in the future. Per capita flour consumption in the United States is highly predictable and, of course, so is population. Accurate flour export projections are available from the Foreign Agriculture Service of the U.S.D.A. These factors together will create very accurate estimates of total U.S. flour pro­duction. Millfeeds are a very predictable percentage of this figure. Millfeed imports and exports are a small percentage of domestic disappearance and can either be disregarded or forecasted, if some basis exists to do so.

It was assumed that millfeed production in Kansas, Nebraska, Iowa, and Missouri will correlate very well with total U.S. production, so an estimate of production inthose four states alone was not made.

Page 20: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

CHAPTER III

NUTRITIVE ANALYSIS

General

Growing animals and birds, and adult animals and birds have differing nutritive requirements at various stages in their lives. These differences are also in effect for different animals and birds at the same stage in life. Broiler chickens grow from baby chicks to adulthood in eight weeks. They must have feeds which are high in energy and protein with very low fiber. Additionally, they must have the right balance of amino acids within the existent protein. The moisture level of the feed must be low enough so that it does not spoil. Feed for milk cows is lower in protein and energy and higher in fiber. Amino acid balance is not as important as in broilers, because reminants can convert protein better than single stomached animals and birds. All feeds have requirements for the various vitamins and minerals which must be observed.

-11-

Page 21: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-12-

The animal nutritionist’s task, in a nutshell, is to formulate feeds for each of the animal and poultry categories his customers are feeding designed to maximize their performance with minimum feed cost. All of this and still maintain profitable conditions for his company!

This task is highly complex. The first step is to design a feed which is ideal for the needs of the customer. Nixon's top selling feed for swine is 40% finisher supplement. It is designed to be fed along with the customer's home grown corn or milo to post weaning age, market-bound hogs. It is the best feed we have found at doing the job and our customers appar­ently agree with us.

40% finisher supplement has several requirements which must be met. These include minimum percentages of protein and fat in the feed plus maximum percentages of moisture and fiber. In addition the right mix of amino acids must be present in the available protein and minimum levels for vitamins and minerals must be present. There are as many as 40 of these requirements which must be met in a given feed. All are important but for our purposes the most important are:

1. Protein (minimum).

Page 22: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-13-

2. Fat (minimum).3 . Fiber (maximum).4. Energy (minimum) for ruminants.5. Productive energy (minimum) for non-ruminants.

16. Moisture (maximum).After the feed is designed, it is subject to

modification at any time. Generally speaking, a well designed feed will be utilized over several years' time, mainly because it performs well and customers keep buying it.

There are a myriad of combinations of feed ingre­dients from which a given feed can be made. The number of potential ingredients is very large. Basically, the combination is determined by price and availability of ingredients. The following are major ingredients:

1. Wheat standard middlings.2. Corn.3. Milo.4. Forty-four per cent soybean meal.5. Cottonseed meal.6. Animal fat.

^Interviews with Tom James, Nutritionist, NCM, July, 1965.

Page 23: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-14-

7. Dehydrated alfalfa pellets.Each can be broken down into its composition of

nutrient contributions. The reader is directed to Table 2 which shows the compositions of the most important nutrient contributions of the major ingredients.

The task of the nutritionist is complicated by the fact that the price, and at times the availability, of ingredients change frequently. The price changes can be quite large. Table 3 provides a nineteen quarter analysis of major feed ingredient prices.

Therefore, the nutritionist examines each of the feeds in his line at frequent intervals to formulate them at the cheapest possible cost, always meeting the nutrient composition requirements. Until six or seven years ago, this work was done only with the aid of the calculator. Today most large feed manufacturers (including Nebraska Consolidated Mills),and some small ones, are utilizing the techniques of linear programming with the aid of the computer.

Table 4 is an example of the form utilized by NCM in setting up the computer run on a given feed.Each ingredient has a code number and each feed a formula number. The cost (expressed in dollars and

Page 24: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-15-

TABLE 2NUTRIENT ANALYSISa

Percentage Calories/lb.

Produc­tive

Protein Fat Fibre Moisture Energy Energy

Wheat Midds 16.50 4.00 6.70 14.00 720 600Corn 8.60 3.90 2.00 15.50 801 1140Milo 8.40 2.90 2.20 15.00 778 110044% Soybean Meal 45.00 0.50 7.00 12.50 796 761

50% Soybean Meal 49.00 0.50 3.00 12.50 801 790

CottonseedMeal 41.00 4.00 13.00 10.00 717 720

Fat • • 100.00 • • • • 2012 2900Alfalfa 17.00 2.00 26.00 10.00 452 300

aFrom interviews with Tom James, Nutritionist, NCM, July, 1965.

Page 25: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-16-

TABLE 3aAVERAGE PRICES BY QUARTER

DateKC

MiddsKCCSM

KCSBM

KCCorn

KCMilo

WeightedAverage

J-M 65 39.90 65.70 70.30 47.10 42.00 57.100-D 64 41.90 68.40 69.20 45.70 41.20 56.50J-S 33.50 66.20 68.70 45.00 42.00 55.90A-J 31.90 62.40 69.20 46.00 40.60 55.50J-M 35.60 71.80 77.50 44.60 39.00 59.100-D 63 42.10 76.60 77.40 43.60 38.60 59.50J-S 38.00 75.60 76.60 46.80 39.60 60.60A-J 32.20 71.80 69.40 44.60 40.00 56.60J-M 39.70 75.10 73.30 44.60 39.80 58.400-D 62 32.40 72.40 72.00 42.10 36.20 56.10J-S 34.10 66.60 73.30 41.10 37.20 55.40A-J 28.90 64.20 66.50 41.10 37.80 52.90J-M 34.80 67.30 62.80 39.30 38.00 51.600-D 61 36.60 66.40 61.30 41.10 36.60 51.30J-S 28.10 63.00 68.90 41.40 38.80 53.80A-J 31.20 62.80 73.00 38.90 36.80 53.90J-M 32.90 59.20 63.40 38.90 34.40 49.700-D 60 34.20 59.80 51.40 36.80 31*. 40 44.60J-S 29.40 61.80 54.50 41.80 33.60 48.00

a„From: "Grain Market News" and "Feed MarketNews" Grain Division, A.M.S., U.S.D.A., Washington, D.C.

Page 26: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965
Page 27: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-18-

and cents per cwt.) of each ingredient is placed beside the ingredient code number. The program contains the contributions of each ingredient. The requirements of the feed formulation are shown at the bottom of the sheet. The computer is fed this information together with the program. It then calculates the least cost com­bination of ingredients. To do this requires a tremendous number of individual calculations such that if it were done by a human with the aid of an automatic calculator,several days would be required to calculate one feed.

2An IBM 1620 will run one feed in 30-90 seconds.The trend to computerized nutrition has changed

the usage pattern of all the major ingredients and will continue to do so as more feed manufacturers utilize this technique. The pre-computer nutritionist used rules of thumb to a great degree in feed formulation. In his mind a given feed contained certain ingredients. These ingre­dients varied within much narrower limits than is the case with the computer. For instance, the nutritionist might have said millfeeds could be as low as 15% or as

2 . .Interviews with Marvin Wilkening, Head Nutrition­ist, NCM, 1965.

Page 28: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-19-

high as 40% in 27% hog finisher supplement. Today'scomputerized nutritionist will allow them to vary from

3.3% to 100%.The computer is much more sensitive to price

changes than its human counterpart. it has no prejudices for or against any ingredient. This being the case any given ingredient's usage pattern will vary much more than previously. Feed usage patterns over the seasons will have less effect on ingredient usage and ingredient price levels will have more effect. This includes price changes of competing ingredients as well as the one in question.

Interviews with Wallace Bierman, Nutritionist, Nixon and Co., July, 1965.

3

Page 29: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-20-

Layer Feeding

In the Kansas City midds marketing area the major poultry feeding activity consists of laying chickens and turkeys for meat. Broilers are not a significant factor.

There is extensive activity in our marketing area in the egg business. There are a few large, vir­tually integrated businesses but also a lot of individual farm families producing eggs on a relatively small scale. It is these families to whom Nixon looks for its laying feed market.

Chicks are normally one day old when bought in the spring. They are placed in brooding houses and fed a chick starter. In the complete form this is 18-20% protein.

When they leave the brooder house and enter the laying house, they are in the pullet stage. They are fed a pullet developer of 16-18% protein.

The last step is the productive layer stage.Here a complete ration of 16-20% protein is fed. Alter­natively the feeder can buy a supplement of 26-36% pro

4tein to be blended with locally produced grains.

James Interview.4

Page 30: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-21-

TABLE 5NIXON POULTRY FEEDS

Allowable AverageMidds Level of

Feed Protein Variance % Midds Used

32% Layer Supplement 3 2% .3-100 13%26% Layer Supplement 26% .3-100 40%20% Complete Layer 20% .3-100 15%16% Complete Layer* 16% .3-30 15%20% Chick Starter 20% .3-20 15%18% Chick Grower 18% .3-30 20%

PERCENTAGEAT

TABLE 6FOR LAYER TYPE FEEDS NIXON, OMAHA

Feed Average Variance

32% Supplement 29% 27-34%26% Supplement 50% 47-55%16-20% Complete 21% 18-25%

Page 31: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

o ** fA -> ?

'A O n ‘J>

x<;

■ ■'X.

-

q-<AVA Ui'• VAc -£t:Gi

Ll

o' O'

Page 32: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-23-

Starter and Grower feeds are very seasonal in nature (see Graph 1) and were not counted.

"Middlings Percentage" weight computation32% .13 X .29 = .03826% .50 X .40 = . 20

16-20% .21 X .15 = .031Weight = .269 = .27

Page 33: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-24-

Turkey Feeding

Nixon does not make turkey feeds. The turkey industry in our area has been characterized by frequent over-production, low prices and credit losses. Today it is almost completely in the hands of large, well- financed firms who provide feed and poults to a grower and pay him for his services. All the market risk is borne by the large firm. This firm is almost always either a processor or a feed manufacturer and is some­times both acting jointly. Nixon has never chosen to participate in it and, therefore, my information is more sketchy than on other lines.

The turkey market is highly seasonal with almost all market birds slaughtered in the fall. The baby birds are placed with the grower in the spring. They are fed in four stages; the pre-starter feed, the starter, grower,

5and finisher. Breeder flocks are carried on a fifth feed.

Ibid.5

Page 34: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-25-

TABLE 7"TYPICAL" PERCENTAGE SOLD DATAPre-Starter F e e d .......... 2%Starter.................... 8%Grower................... 30%Finisher................. 60%Breeder................... 10%

TABLE 8

"TYPICAL" TURKEY FEEDS

Feed ProteinEnergy

Cal./lb."Average" Level of Midds Used

Pre-Starter 24-30% 875 3%Starter 24% 875 3%Grower 18-24% 900 1 0%Finisher 16-18% 975 17%Breeder 17% 900 28%

"Middlings Percentage" weight computationPre-Starter . 0 2 X .03 = . 0 0 1

Starter .08 X .03 = . 0 0 2

Grower . 30 X . 1 0 = .030Finisher .60 X .17 = . 1 0 2

Breeder . 1 0 X .28 = .028Weight .163 .16

Page 35: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-26-

Hog FeedingAs in the case of all feeds there are supplements

and complete feeds made in the hog line. Supplements are designed to be used in conjunction with farm grown grains. In the area of our interest this is almost always either corn or milo, usually corn. Supplements are, therefore, higher in protein content and lower in energy than com­plete feeds. *

The hog line of feeds usually consists of starter and finisher feeds for growing pigs and sow feed for breed­ing stock.

Starter feeds supplement the sow's milk in the early stages and prepare the baby pig to go on finisher rations shortly after weaning. They represent small volume in the total intake during a market hog's lifetime. They usually are 18-22% protein and are almost always sold as complete feed.

Finisher feeds carry the growing pig from 40-50 pounds, 6 weeks old to market weight. A growing pig will consume approximately one-fourth ton of them in its life­time. They represent the largest volume item in the hog line. Protein content ranges from 12-18% in the complete

Page 36: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-27-

feeds. They are high in energy and therefore invariably utilize a high percentage of corn or milo.

Many hog men in our area buy supplements for feeding in conjunction with locally grown grains. These supplements range from 27-40% in protein content. The choice of protein will depend on the feeders' own ideas and also whether they are on pasture, feedlot, or in confinement. Pigs on prime pasture require less feed and higher protein.

Feeds for breeding stock can be complete or asupplement. In the supplement form they will average30% protein. In the complete form the average is 12%

6protein.

TABLE 9PERCENTAGE OF EACH TYPE FEED SOLD AT NIXON, OMAHA

Feed Average Variance

35-40% Supplement 49% 32-63%27% Supplement 22% 15-31%12-16% Grower 1 0% 14-16%Starters 14% 8-22%Sow Feeds 5% 3-9%

6Ibid.

Page 37: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-28-

TABLE 10NIXON SWINE FEEDS

Feed ProteinAllowable

Midds Variance"Average" Level of Midds Used

40% Supplement 40% .3-100% 3%27% Supplement 27% .3-100% 25%16% Grower 16% 2-1 0 0% 13%14% Grower 14% .3-100% 13%12% Grower 1 2% 2-50% 13%18% Starter 18% .3-100% 20%22% Starter 22% .3-100% 7%30% Sow Supplement 30% .3-100% 7%

"Middlings Percentage" weight computation35-40% .49 X .03 = .01527% . 22 X .25 = .05512-16% . 1 0 X .13 = .013Starter .14 X .14 = . 0 2 0

Sow .05 X .07 = .004Weight .107 .11

Page 38: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-29-

Cattle Feeding

The production of beef in our area is almost always done in two stages, ranching and feeding. The semi-arid regions which occupy a great deal of the Nebraska-Kansas territory are utilized to a great extent as pasture for brood cows producing calves. The heavy corn and milo producing areas of our four state area are heavily populated with yearling steers and heifers in the process of being fattened on high-grain rations for market. These two stages are distinctly different in their feeding requirements.

The rancher is desirous of producing yearlings for sale to feeders at the lowest possible cost. He has lots of grassland at his disposal which he hopes will be the only source of feed he will need. However, he is very rarely able to completely do this because of the vagaries of nature. He must buy feeds for his cows, bulls and calves at certain times to supplement grazing. This normally happens in the winter. Depending on the quantity and quality of grass available, the rancher will use feeds ranging from 10-40% in protein content.

The feeder has plentiful supplies of high energy low protein feed grains available. He needs a high

Page 39: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-30-

protein supplement for blending with his grain. This can vary from 20-41% protein but is heavily concentrated in the higher protein range. In rare instances lower protein complete feeds of 10-18% are fed.

In the past seven years the use of syntheticurea as a source of protein for ruminant animals (cattleand sheep) has doubled in the U. S. This is replacinghigh protein soybean and cottonseed meal. Urea rarelychanges in price whereas the natural protein ingredientsare very volatile in their price fluctuations. In recentyears urea based feeds have been lower priced than"natural" protein feeds of the same description. Therehas been a tendency for feeders not to use urea basedfeeds because of a belief that it "burns" the animal andcauses high death losses. This belief is being dispelledand the popularity of urea based feeds is increasing. In1956 the U. S. used 60,000 tons of urea for animal feedwhereas in 1963 the usage was 130,000 tons. Nixon's

7figures bear this out.This is significant for millfeeds as they combine

very well with urea to make an excellent supplement. The

7"Usage of Urea Doubles in Seven Years," The Southwestern Miller, September 8 , 1964, page 54.

Page 40: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-31-

reader will note that a 41% urea based supplement utilizes a much higher percentage of middlings than does a 40% natural protein supplement.

Urea based was 17% of total beef feed in last observed quarter.

For comparison of Nixon natural versus urea based tonnage, please see Graph 2.

"Middlings Percentage" weight computation:1964 tonnage was arbitrarily taken rather than average over four years because of the strong urea trend. In that year natural feeds had 84% of total cattle feeds sold and urea based had 16%.

40% Natural .36 X .84 X . 0 0 = . 0 0 0

32% Natural .48 X .84 X .15 .06018-22% . 1 1 X .84 X . 20 = .0181 0 -1 2% .05 X .84 X .15 = .00641% Urea .50 X .16 X .25 = . 0 0 2

32% Urea .50 X .16 X .35 = .003Weight .089 .09

Page 41: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

_/vVo >

\ * V1 r> /

i r v ?

> vS n /

'< < t ?

-ft*

H ?

1* '*! -

J-1 *=»'

.^o_i£*t

- §

Page 42: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

TABLE 11NIXON BEEF FEEDS

-33-

FeedAllowableMidds

Protein Variance"Average" Level of

Midds Used

40% Supplement (Nat.) 40% • • •

41% Supplement (Urea) 41% 8-40% 25%32% Supplement (Nat.) 3 2% 5-30% 15%32% Supplement (Urea) 32% 10-60% 35%22% Complete (Nat.) 22% .3-100% 35%20% Complete (Nat.) 20% 2-8% 5%18% Complete (Nat.) 18% .3-100% 25%12% Complete (Nat.) 1 2% 5-30% 25%10% Complete (Nat.) 1 0% • • •

TABLE 1 2

PERCENTAGE OF TYPE FEED SOLD AT NIXON, OMAHA

Feed Average Variance

Natural Based

40% Supplement 36% 28-48%32% Supplement 48% 38-59%18-22% Complete 1 1 % 8-17%10-12% Complete 5% 2-1 0%

Urea Based

41% Supplement 32% Supplement

50%50%

17-76%24-83%

Page 43: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-34-

Dairy Feeding

Our four state area is not heavily populated with dairy cattle to the extent that Wisconsin and Minnesota are. Dairying is a major factor, however, in the feed business and must be considered.

The dairyman, like the rancher, is desirous of growing all of his own feed on the home farm. He tradi­tionally falls short of his mark and must supplement with a commercially manufactured feed. He raises cattle from birth to death and the large bulk of his feed is for the milk producing cow. His need is for a complete feed of 15-20% protein with plenty of sugar present to make it more palatable. He can also utilize a 30-35% protein concentrate along with home grown grains.

It is interesting to note that the dairyman was one of the first animal feeders to utilize wheat bran as a supplemental feed. He would have it mixed at the local mill with his own oats, barley and/or corn as a source of protein. He might also buy bran in bags and sprinkle it on top of ground grain and ensilage at milking time. This usage continues but to a much smaller extent and is probably on the way out completely. The sacked bran business in our territory is almost negligible now.

Page 44: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-35-

Nixon has never had very much dairy feed tonnage.The small amount they do have has shifted from 80% of 30%protein supplement in 1960 to 60% of 16% complete today.Those are the only two dairy feeds in their line. Thisis not unusual for a dairy manufacturer, however, andgives him the opportunity to reduce costs with long runs

8on the same feed.TABLE 13

NIXON DAIRY FEEDS

Allowable "Average" LevelFeed Protein Midds Variance of Midds Used

30% Supplement 30% .3-25% 1 0%16% Complete 16% .3-25% 15%

Note:Each is 50% of the volume."Middlings Percentage" weight computation

30% .50 x .10 - .0516% .50 x .15 = .08

Weight = .13

8James interviews.

Page 45: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-36-

Conclusion

The foregoing material on the usage of middlings in the various animal and poultry rations was designed to develop the weighting process in Chapter V and to give the reader background information on animal and poultry feeding.

Page 46: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

CHAPTER IV

COMPETING INGREDIENTS

The four major competitors of middlings for in­clusion in animal feed formulae are corn, soybean meal, milo and cottonseed meal. As in the case of millfeed, there are monthly average prices of daily quotations for these items in Grain Market News and Feed Market News. These figures were utilized and further condensed into quarterly averages for purposes of generating the variable, competing ingredient prices.

The percentage of each of these ingredients utilized in various feed formulae varies as their relative prices change. All of them are usually present in most feeds in our area, however. Strictly on the basis of my general knowledge of their relative importance, I weighted them as follows for purposes of the multiple correlation.

M i l o ................... 16 2/3%Cottonseed Meal . . . . 16 2/3%C o r n ................... 33 1/3%44% Soybean Meal . . . . 33 1/3%

37-

Page 47: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965
Page 48: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

G-R

flfU

- ̂5-

vA >■•i •b y•5f 7*

A /6

!* «

v/\ M£ J- /*7 -J

;̂-*r

£ J/

J\ o_

o v: -■>< V

vA 7.

v So M P7

Page 49: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

CHAPTER V

WEIGHTING ANIMAL NUMBERS

An analysis of the feed tonnage figures issued by the American Feed Manufacturers Association compared with actual animal inventory figures indicates that there are very pronounced seasonal patterns in feed usage. The reader is invited to examine Table 14 to see this in detail.

Furthermore, it is readily apparent that a cow will consume several times the volume of feed of a chicken in a given time period. Therefore, several weighting factors were applied to the animal numbers to make them more meaningful for purposes of the correlation analysis. Four weighting factors will be applied to each quarter's animal inventory.

1. This factor will be termed the "absoluteweight" for each animal type and stems from their varying capacity to consume feed. Hogs were selected as the standard animal category because variance in feed tonnage for hogs

40-

Page 50: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-41-

A.F.M.A. FEED TONNAGE 000'S OMITTED9

TABLE 14

Date Layer % Turkey % Dairy % Hog % Cattle % Total

J-M 97 14 1 2 2 49 7 274 40 240 37 672O-D 64 83 1 2 23 3 48 7 291 42 252 36 697J-S 81 14 62 1 1 41 7 286 48 118 20 588A-J 104 18 53 9 40 7 277 47 119 20 593J-M 85 14 1 1 2 46 7 266 43 209 34 6170-D 63 81 13 14 2 46 7 272 42 222 36 635J-S 84 14 48 8 40 7 294 49 128 21 594A-J 1 1 2 18 42 7 40 7 281 46 134 22 609J-M 89 14 1 0 2 48 7 261 40 237 37 6450-D 62 75 1 2 10 2 43 7 262 42 225 37 615J-S 91 16 44 8 36 7 271 49 106 20 548A-J 113 21 31 6 39 7 248 46 1 1 2 20 543J-M 98 15 1 2 5 50 8 235 40 196 33 5870-D 61 97 16 24 4 51 8 239 39 205 33 616J-S 99 18 71 13 40 7 236 44 93 17 539A-J 140 23 82 13 42 7 247 40 109 18 620J-M 105 19 17 3 45 8 208 37 181 32 5560-D 60 94 17 2 1 4 45 8 199 37 181 33 540J-S 1 0 1 2 1 56 1 1 39 8 198 40 93 19 487

Total........................................ 11,301M e a l ........................................ 594

aFrom: Confidential sales analysis of feed tonnagepublished by the American Feed Manufacturers Association, Chicago 4, Illinois, each quarter.

Page 51: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-42-

TABLE 14— Continued

Average Quarterly Percentages

Quarter Layer Turkey Dairy Hog Cattle

4 14 3 7 40 353 17 10 7 46 192 20 9 7 45 20

1 15 3 7 40 35

Total 66 25 28 171 109Mean 16 6 7 43 27

Animal Weights for Given Quarter

4 87 50 1 0 0 93 1303 106 166 1 0 0 107 702 125 150 1 0 0 106 741 94 50 1 0 0 93 130

Page 52: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-43-

had the highest correlation with variance in animal numbers.The actual mean animal numbers were then computed for each category. (The mean number of hogs on farms for the 19 periods was 22.4.) A simple ratio and proportion was set up stating that the ratio of hog numbers to percentage of the total feed accounted for by hogs was 22.4/43. 22.4/43equals .52, the factor to be utilized for weighting each animal category in terms of hog feed consumed equivalent.The mean number of layers was 39.6. They accounted for 16% of the total feed consumed .52 = x/16 - x = 8.3 8.3 39.6 - .21, the"absolute weight" to be used for layers. Similarly the "absolute weight" for cattle i .72, for turkeys is .20 and for dairy is .58

Recap of Absolute Weights Layers .21Turkeys .20Dairy .58Cattle .72Hogs 1.00

Page 53: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-44-

2. Cattle obtain a larger proportion of theirfeed requirements from pasture in Quarters 2 and 3 than they do in 1 a»d 4. This phenomenon gives rise to a factor termed "quarter weight by animal type." It was generated by analyz­ing the A.F.M.A. feed tonnage records by animal types and by quarters. For instance, for Quarter 1-65 it was determined that layers consumed 14% of the total feed, turkeys 2%, dairy 7%, hogs 40% and other cattle 37%.A similar analysis was made for each quarter over a 19 quarter period.It developed that for Quarter 4 layers had 12%, 12%, 13%, 16% and 17% respectively for each of the five years. The mean was 14%.Each quarter/animal was similarly analyzed.The means for each quarter/animal were th6n totaled and a mean generated for each. The results were that layers accounted for 16% of the feed, turkeys 6%, dairy 7%, hogs 43% and other cattle 27%. The "quarter weight by animal type" was obtained by dividing the mean for the quarter by the mean for the year.The results are as follows;

Page 54: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-45-

INVENTORIES OF HOGS ON FARMS 000'S OMITTED3

TABLE 15

Date Kansas Nebraska Missouri Iowa Total

3/65 1,304 2,410 3,500 11,652 18,86612/64 1,407 2,929 4,203 13,829 22,3689/64 1,480 2,978 4,084 14,813 23,3556/64 1,403 3,039 4,025 14,358 22,8253/64 1, 275 2,384 3,804 11,648 19,11112/63 1,421 3,116 4,568 14,557 23,6629/63 1, 575 3,273 4,538 15,271 24,6576/63 1,576 3, 377 4,193 14,802 23,9483/63 1,371 2,509 4,047 12,261 20,18812/62 1,496 3,055 4,586 14,413 23,5509/62 1,514 3,178 4,449 14,972 24,1136/62 1,530 3,311 4,151 14,655 23,6473/62 1,371 2, 509 4,047 12,261 20,18812/61 1,411 3,121 4,437 14,130 23,0999/61 1,402 3, 210 4,319 15,123 24,0546/61 1,430 3,218 4,070 15,250 23,9683/61 1,186 2,450 3,818 12,017 19,47112/60 1,292 2,871 4,493 14,150 22,8069/60 1,149 2,720 3,962 14,267 22,098

Total • • • • • . 427,974Mean • • • • • • • • 22,420

aFrom; Kansas Crop and Livestock Reporting ServiceU.S.D.A. Statistical Reporting Service and Kansas StateBoard of Agriculture , Weekly Crop Bulletins, 1960-1965.

Page 55: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

46-

TABLE 16TURKEYS PRODUCED 000'S OMITTED3

Year Kansas Nebraska Missouri Iowa Total

1965 712 1,126 8,062 7, 395 17,2951964 816 1,139 6 , 735 7,985 16,6751963 760 913 5,221 7,964 14,8581962 834 1,151 4,103 7,792 13,8801961 1,180 1,475 5,699 9,836 18,1901960 863 1,113 4, 285 7,675 13,936

Total • . • • • 94,834Mean • • • • ’ • • 15,806

aIbid.

Page 56: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-47-

TABLE 17INVENTORIES OF OTHER CATTLE ON FARMS

000'S OMITTED3

Date Kansas Nebraska Missouri Iowa Total

1/1/65 4,848 5,701 3,827 6 , 503 20,8791/1/64 5,104 5,744 3,758 6 , 263 20,8691/1/63 4,867 5,457 3,633 6 , 380 20,3371/1/62 4,599 5,094 3,547 5,729 18,9691/1/61 4,176 4,784 3,390 5,526 17,8761/1/60 4,039 4, 716 3,257 5,717 17,729

Total , , 116,659Mean 19,443

aIbid.

Page 57: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-48

TABLE 18INVENTORIES OF MILK COWS ON FARMS

(2 YRS & OLDER)000’S OMITTED&

Date Kansas Nebraska Missouri Iowa Total

1/1/65 311 301 608 835 2,0551/1/64 327 304 633 861 2,1251/1/63 355 316 673 888 2,2321/1/62 374 340 716 925 2,3551/1/61 386 350 709 934 2,3791/1/60 390 356 723 943 2,412

Total • - • • • • 13,558Mean • • • • - • • • 2,260

a , . , Ibid.

Page 58: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-49-

INVENTORIES OF LAYERS ON HAND AVERAGE DURING MONTH

OOO'S OMITTED3

TABLE 19

Date Kansas Nebraska Missouri Iowa Total

3/65 4,888 6 , 686 6,312 17,316 35,20212/64 4,872 6,858 6 , 386 17,352 35,4689/64 4,410 6,091 6 , 240 16,506 36,2476/64 4,063 5, 992 6 , 290 16,198 32,5433/64 4,588 6,778 7, 266 17,398 35,97012/63 4, 910 7,158 7, 213 17,890 37,1719/63 4,560 6,624 6,918 16,400 34,5026/63 4,474 6,480 7,156 17,192 35,3023/63 4, 969 7, 304 7,889 19,172 39,33412/62 5,550 7,616 8,519 19,824 41,5099/62 5,162 7,009 7,932 18,500 38,6036/62 5,043 7, 204 8,278 19,226 39,7513/62 5,582 8,114 9,058 21,532 44,28612/61 6,156 9,039 9,696 22,795 47,6869/61 5,476 7,795 8,138 20,214 41,6236/61 5,474 7,652 7, 997 20,436 41,5593/61 6,029 8,582 8,638 22,613 45,86212/60 6 , 676 9, 286 9, 228 2 2 , 1 2 0 47,3109/60 6,146 8,320 8 , 350 19,360 42,176

Total • • • • • • • 751,702Mean • • • • • • • • 39,563

aIbid.

Page 59: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-50-

Quarter Layers Turkey Dairy Hog Cattle4 .87 .50 1 . 0 0 .93 1.303 1.06 1 . 6 6 1 . 0 0 1.07 . 702 1.25 1.50 1 . 0 0 1.06 .741 .94 .50 1 . 0 0 .93 1.30

3. A "weight for each quarter" was also computed. Total feed tonnage for the 19 quarters was obtained and the mean per quarter computed. Then actual tonnage for the quarter was taken. Please see Table 20 for the actual figures.The actual tonnage for the quarter was divided by the mean to obtain a weight for each of the 19 quarters. These weights were then grouped by Quarters 1-4 and a mean computed for each quarter. Here is the "weight for each quar­ter."Quarter 1 2 3 4Weight 1.03 .99 .93 1.04

4. Feed tonnage has played a very important role in the weighting method so far. It was there­fore decided to weight the relative importance of middlings in each type feed. A descriptive

Page 60: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-51-

FEED TONNAGE, QUARTER WEIGHTING MEAN = 594a

PercentQuarter of Mean

1- 6 5 ................................ 1134 - 6 4 ................................ 1173- 6 4 ................................. 992- 6 4 ................................ 1001- 6 4 ................................ 1034- 6 3 ........ 1073- 6 3 ................................ 1002- 6 3 ................................ 1021- 6 3 ................................ 1084- 6 2 ................................ 1033- 6 2 ................................. 922- 6 2 ................................. 911- 6 2 ................................. 994- 6 1 ................................ 1043- 6 1 ................................. 912- 6 1 ................................ 1041-61................................ 944- 6 0 ................................. 913- 6 0 ................................. 82

RECAP BY QUARTERS

TABLE 20

1 2 3 4

1965 113 -- -- --1964 103 100 99 1171963 108 102 100 1071962 99 91 92 1031961 94 104 91 1041960 -- -- 82 91

Total 517 397 464 522Mean 103 99 93 104

aSee Item 3, Chapter V.

Page 61: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-52-

section on this subject preceded in Chapter III. It is the basis for these weights.They were computed by analyzing the percen­tages of the type feed sold compared with the "average" percentage of midds used in that feed. The "middlings percentage" weights are as follows:

Layers . 27Turkey .16Dairy .13Hog . 1 1

Cattle .09There are two weights to be applied to each animal

type and two weights to be applied to each quarter. They

can be summarized as follows:Quarter 1 2 3 4Layers .05 .06 .07 .06

Turkeys . 0 2 .05 .04 . 0 2

Dairy .08 .07 .07 .08Cattle . 1 2 .06 .06 . 1 2

Hogs COo• .08 .08 .08These weights were then applied to total animal

inventories in Tables 15-19 to secure the "Weighted Animal and Poultry Inventory Numbers" in Table 21.

Page 62: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-53-

TABLE 21WEIGHTED ANIMAL AND POULTRY NUMBERS3

Quarter Layers Turkeys Dairy Cattle Hogs Total

1-65 1.74 .35 .17 2.51 1.51 6.284-64 2.13 .33 .17 2.51 1.79 6.933-64 2.53 .67 .15 1.25 1.87 6.472-64 1.95 .84 .15 1.25 1.82 6 . 0 11-64 1.80 .33 .17 2.51 1.53 6.344-63 2.23 . 30 .18 2.44 1.90 7.053-63 2.42 .60 .15 1 . 2 2 1.98 6.372-63 2 . 1 2 . 75 .15 1 . 2 2 1.91 6.151-63 1.97 . 30 .18 2.44 1.62 6.514-62 2.49 . 28 .19 2,28 1.89 7.133-62 2.70 .56 .17 1.14 1.93 6.502-62 2.39 .70 .17 1.14 1.89 6.291-62 2 . 2 2 .28 .19 2.28 1.62 6.594-61 2 . 8 6 .36 .19 2.15 1.85 7.413-61 2.91 .73 .17 1.07 1.93 6.812-61 2.50 .91 .17 1.07 1.92 6.571-61 2.30 .36 .19 2.15 * 1.56 6.564-60 2.84 . 28 .19 2 . 1 2 1.82 7.253-60 2.95 .56 .17 1.06 1.77 6.51

aFrom Tables 15-19 and Item 4, Chapter V.

Page 63: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-54-

The forecaster needs to continually re-examine and re-evaluate the subject of weighting this data as there are trends taking place which change the averages as the years pass by. The growth of beef cattle numbers and the decline of layers are examples.

Many judgment factors need to be made in this process eo the forecaster must be aware of the entire agricultural spectrum and the developments within it.The very process of gathering and evaluating the figures shown here has opened the writer's eyes to several things he had not known of before.

There are many more figures available on animal and poultry numbers, livestock movements, feed grain pro­duction, etc., which were examined but not used. Those which were used were felt to be the most important.

Page 64: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

CHAPTER VI

THE CORRELATION ANALYSIS

It is now possible to perform the correlationanalysis thus testing the hypothesis. The method used isprescribed in Spencer, Clark and Hoguet, Business and

1Economic Forecasting, pages 45-63. One table (Table 22, The Correlation Analysis) and several graphs (Graphs 5-9) are necessary to do this.

The goal is to predict the change in the dependent variable, Y (millfeed price), by predicting the changes in the independent variables, X]_, X2 , and X3 .

The first step is to graph each of the independent variables versus the dependent variable. This is done for animal numbers, X^, in Graph 6 ; for domestic disap­pearance of millfeeds, X2/ in Graph 5? and for competing ingredients, X^, in Graph 7. The one which appears to have the highest correlation is selected as X^, etc.

■^Milton Spencer, Colin Clark, and Peter Hoguet, Business and Economic Forecasting, (Homewood, 111.,R. D. Irwin Co., 1961), p.45-63.

-55-

Page 65: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-56-

THE CORRELATION ANALYSISTABLE 22

ObservationPeriod(Quarters)

(1)

K.C. Midds Price Y

(2)

WeightedAnimalNumbers

X1(3)

DomesticDisappearance

X2

(4)

CompetingIngredients

X3

(5)

1-65 39.90 6.28 1,068 57.104-64 41.90 6.93 1,283 56.503-64 33.50 6.47 1,091 55.902-64 31.90 6.01 1,365 55.501-64 35.60 6.34 1, 218 59.104-63 42.10 7.05 1,326 59.503-63 38.00 6.37 1,233 60.602-63 32.20 6.15 1,193 56.601-63 39.70 6.51 1,217 58.404-62 32.40 7.13 1,252 56.103-62 34.10 6.50 1,211 55.402-62 28.90 6.29 1,146 52.901-62 34.80 6.59 1, 283 51.604-61 36.60 7.41 1,292 51.303-61 28.10 6.81 1,247 53.802-61 31.20 6.57 1,115 53.901-61 32.90 6.56 1,244 49.704-60 34.20 7.25 1,298 44.603-60 29.40 6.51 1,207 48.00

Total 657.40 125.73 23,289 1,036.50Mean 34.60 6.62 1,226 54.60

Y X1 x 2 X3P = - 4 2 . 4 0 + 11.54X-L - . 0 2 6 X 2 + . 58X 3

2R = .53

Page 66: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-57-

TABLE 22— Continued

Regression Estimate Chart B

Yb

Residual No. 1

Chart B 2 - 6

Regression Estimate Chart C

Yc

Residual No. 2

Chart C 7 - 8

Regression Estimate Chart D

Yd

Residual No. 3 Chart D 9 - 1 0

Calculated Price P

6 + 8 + 1 0Ratio, Actual Price -1- Cal­culated Price

2 - 1 2(6) (7) (8) (9) (10) (ID (12) (13)

3 2 . 1 0 7 . 8 0 3 . 5 0 4 . 3 0 - . 4 0 4 . 7 0 3 5 j 2 0 1 . 1 33 9 . 6 0 2 . 3 0 - 2 . 1 0 4 . 4 0 - . 7 0 5 . 1 0 3 6 . 8 0 1 . 1 43 4 . 3 0 - ' . 8 0 2 . 9 0 - 3 . 7 0 - 1 . 1 0 - 2 . 6 0 3 6 . 1 0 . 9 32 9 . 0 0 2 . 9 0 - 4 . 2 0 7 . 1 0 - 1 . 3 0 8 . 4 0 2 3 . 5 0 1 . 3 63 2 . 8 0 2 . 8 0 - . 4 0 3 . 2 0 . 8 0 2 . 4 0 3 3 . 2 0 1 . 0 74 1 . 0 0 1 . 1 0 - 3 . 2 0 4 . 3 0 1 . 0 0 2 . 3 0 3 8 . 8 0 1 . 0 93 3 . 1 0 4 . 9 0 - . 8 0 5 . 7 0 1 . 6 0 4 . 1 0 3 3 . 9 0 1 . 1 23 0 . 6 0 1 . 6 0 . 3 0 1 . 3 0 - . 7 0 2 . 0 0 3 0 . 2 0 1 . 0 73 4 . 7 0 5 . 0 0 - . 3 0 5 . 3 0 . 4 0 4 . 9 0 3 4 . 8 0 1 . 1 44 1 . 9 0 - 9 . 5 0 - 1 . 3 0 - 8 . 2 0 - 1 . 0 0 - 7 . 2 0 3 9 . 6 0 . 8.23 4 . 6 0 - . 5 0 - . 2 0 - . 3 0 - 1 . 4 0 1 . 1 0 3 3 . 0 0 1 . 0 33 2 . 2 0 - 3 . 3 0 1 . 5 0 - 4 . 8 0 - 2 . 8 0 - 2 . 0 0 3 0 . 9 0 . 9 43 5 . 6 0 - . 8 0 - 2 . 1 0 1 . 3 0 - 3 . 6 0 4 . 9 0 2 9 . 9 0 1 . 1 64 5 . 1 0 - 8 . 5 0 - 2 . 3 0 - 6 . 2 0 - 3 . 7 0 - 2 . 5 0 3 9 . 1 0 . 943 8 . 2 0 - 1 0 . 1 0 - 1 . 1 0 - 9 . 0 0 - 2 . 3 0 - 6 . 7 0 3 4 . 8 0 . 8 13 5 . 4 0 - 4 . 2 0 2 . 3 0 - 6 . 5 0 - 2 . 2 0 - 4 . 3 0 3 5 . 5 0 . 8 83 5 . 3 0 - 2 . 4 0 - 1 . 0 0 - 1 . 4 0 - 4 . 7 0 3 . 3 0 2 9 . 6 0 . 9 04 3 . 3 0 - 9 . 1 0 - 2 . 5 0 - 6 . 6 0 - 7 . 6 0 1 . 0 0 3 3 . 2 0 1 . 0 33 4 . 7 0 - 5 . 3 0 - . 1 0 - 5 . 2 0 - 5 . 7 0 . 5 0 2 8 . 9 0 1 . 0 2

Page 67: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

* K*- VaSC

<c -V- ^

r̂ »

<v*Sj •cv. <■..

Page 68: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965
Page 69: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965
Page 70: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

; ~ ~ I

• •

• -

- y ^

:....

..__

......

....

.......

......

......

...;__

Com g

s rrc

Oo

t,\o

o

i, t00

{i/o

o

OI&

AfS

f/X

fa

Page 71: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965
Page 72: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-63-

Columns 1-5 can then be filled in from existingdata.

Chart B depicts millfeed prices on the Y axis and weighted animal and poultry numbers on the X axis.Each of the points is plotted from Tables 21 and 3. It is then necessary to plot a straight line which best"fits" the scatter of points on the graph. This depicts

*the average relationship between the X variable and the Y variable. Generally speaking the higher the animal and poultry numbers, the higher the millfeeds price, and vice versa. From this line can be computed the

y 2 - v i

equation Y^ = y^ + X 2 - x]_ (xl“ x]_) * Yb =11.54X1 - 40.40. Column 6 is completed by substituting for X1 the actual values of X.̂ for each of the periods. Column 7 is Column 2 minus Column 6 and indicates how closely changes in Y can be predicted by changes in X-̂ only.

Chart C depicts domestic disappearance, X2» on the X axis and residual number 1 (Column 7) on the Y axis. Its purpose is to represent the average influence of supply on price after allowance has already been made for the influence of animal numbers on price. The straight line again depicts the average relationship and from it

Page 73: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-64-

the equation Y = 31.30 - . 026X is derived. This equa- tion and Column 4 are used to derive Column 8.

Columns 9-11 are derived in the same manner as described above.

The equation from Chart D is = .58X^ - 33.30. These three equations are combined to give the generalforecasting equation for millfeed price, p = -42.40 +

11.54X1 - .026X2 + .58X3. If the actual values of X9, and X were substituted into that equation for each quarter, the values in Column 12 would be obtained. Or Column 12 can be obtained by adding Columns 6 plus 8 plus 10.

Column 13 compares the forecasted price in Column 12 with the actual price in Column 2. It is expressed as the ratio of Column 2/Column 12. It indicates the accuracy of the prediction. perfect prediction accuarcy is 1.00. Another indication of prediction accuracy is Chart D. Had all the plotted points fallen on the line, the prediction accuracy would have been perfect. A thirdindication is gained by computing the "estimate of deter-

2mination", R . A value of 0 for this means there is nocorrelation at all while 1.0 would indicate perfect cor-

2relation. The value obtained was R = .53.

Page 74: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-65-

None of these indications gave the assurance that the P equation derived above was a good one forpredicting. In order for this equation to be of value

2an R of at least .8 would be needed or an equation that would always predict the price within $2 per ton. This one was off as much as $8.40 per ton in one instance.In order to prove the hypothesis a great deal more work<•would be necessary.

Further work toward improving animal number weighting techniques might help. It might help to study in greater detail the weighting of competing ingredients prices. It is possible that this technique should be utilized for all of the U. S. rather than just Iowa, Nebraska, Missouri and Kansas. Then a study of the differentials between the major markets, Buffalo, Kansas City, Minneapolis and Chicago could be utilized toachieve a forecast for each of them.

Page 75: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

BIBLIOGRAPHY

Crop and Livestock News 1960-65, Iowa Crop and Livestock Reporting Service, Iowa Bldg. Room 50b, Des Moines, Iowa.

Feed Market News, Weekly Summary and Statistics, 1961-65, Grain Division, A.M.S., U.S.D.A., Washington D.C.

*Grain Market News, Weekly Summary and Statistics, 1961-

65, Grain Division, A.M.S., U.S.D.A., Washington, D.C.

Kansas Crop and Livestock Reporting Service, U.S.D.A.Statistical Reporting Service and Kansas State Board of Agriculture, Weekly Crop Bulletins 1960-65.

Kansas State Board of Agriculture, Farm Facts, 1961-62, 1962-63, 1963-64, 1964-65.

Livestock, Meat, Wool Market News, Weekly Summary and Statistics, 1961-65, Livestock Division, Agri­cultural Marketing Service, U.S.D.A., Washington, D.C.

Millfeed, By-Product, Co-Product or Product? Reprint of panel discussion presented at the 61st Annual Convention of the Millers National Federation,May 14, 1963. Available from the M.N.F. Office, 309 W. Jackson Blvd., Chicago, 111.

Millfeed Production and Marketing Practices in Kansas and Missouri, Agri-Research, Inc., Manhattan, Kansas, March, 1965.

-66-

Page 76: FORECASTING MILLFEED PRICES MBA 299 SEPTEMBER 7, 1965

-67-

Nebraska Department of Agriculture and Inspection,Nebraska Agricultural Statistics, Annual Reports 1961, 1962, 1963.

Nixon and Company. Personal interviews with Wallace Bierman, Nutritionist; Dr. Marvin Wilkening,Head Nutritionist; Tom James, Nutritionist; and Russell Kendall, Nutritionist. July and August, 1965.

Product Manual, Nixon Feeds, Omaha, Nebraska.Sales Tonnage Analysis, all available quarters 1960-65,

The American Feed Manufacturers Association, Ch-Lcago 4, Illinois. CONFIDENTIAL.

Spencer, Clark & Hoguet, Business and Economic Fore­casting, Homewood, 111., R. D. Irwin Co., 1961.

The State Department of Agriculture, State of Missouri, Don V. Thomason, Commissioner, The Bulletin, Missouri Annual Crop and Livestock Production by Counties, 1941-60.

U. S. Department of Ag. Livestock and Meat Statistics,1961, 1962, 1963, 1964. Agricultural Marketing Service, Statistical Reporting Service, Economic Research Service, Washington, D.Ci, 1961, 1962, 1963, 1964.

U. S. Department of Ag. Agriculture Statistics, 1961,1962, 1963, 1964. U. S. Government Printing Office, 1961, 1962, 1963, 1964.

"Usage of Urea Doubles in Seven Years," Unsigned Article in "The Southwestern Miller," September 6, 1964, The Sosland Press, Kansas City, Missouri.