Ethridge, D. Development of A Price Reporting System For ...

4
Reprinted from: Proceedings of the Beltwide Cotton Production Conference (1990):37-39. Ethridge, D. "Development of A Price Reporting System For HVI Grading." Beltwide Cotton Conference, 1990.

Transcript of Ethridge, D. Development of A Price Reporting System For ...

Page 1: Ethridge, D. Development of A Price Reporting System For ...

Reprinted from:

Proceedings of the Beltwide Cotton Production Conference

(1990):37-39.

Ethridge, D. "Development of A Price Reporting System For HVI Grading." Beltwide Cotton Conference, 1990.

Page 2: Ethridge, D. Development of A Price Reporting System For ...

37

DEVELOPMENT OF

Professor,

A PRICE REPORTING SYSTEM FOR HVI GRADING Don Ethridge

Depart •• nt ot Aqricultural Econo.ies Texas Tech Un~versity

Lubbock. TX

Abstract

A joint project involving Texas Tech University and the Market News Branch. Cotton Division, Agricultural Marketing Service, USDA is underway to develop a system to identify and report cotton prices and quality pre~lums and discounts under the HVI grading syste.. Develop~ent and testing of .odels and identification of data needs and sources has been accomplished. Selection of systems for i ntensive testing and evaluation in the Texas­Oklaho.a market is in progress. final system selection and adaptation of the system to other markets remains.

Page 3: Ethridge, D. Development of A Price Reporting System For ...

Introduction

After 1I0re than a decade of development, study, evaluation, and debate, the cotton industry has made the decision to adopt high volume instrument (HVI) evaluation ot cotton fiber quality attributes (gra~l.ng) as the industry standard. while there are ObVIOUS problems associated with adopting new systems, ,s tandards, and technology, particularly those as sweepl.ng as HVI, the industry' s decision to implement HVI ,in 1991 suggests a consensus among the leadership that l.ts time has come.

As awareness and understanding of the HVI system and its posdbilities have progressed, there hl!'s been a lagged but growing recognition that the benefl.ts of ~I can be fully realized by the industry as a whole only l.f there is accurate, reliable price information on the fiber quality attributes which HVI generates. , Put another way , all participants in the market need tImely, reliable infor.ation on both the general level of ,market prices and the premiums and discounts associated wl.th the set ot relevant fiber quality attributes.

In response to this recognition, work is in progre~s to i_prove that price information. The purpose of thl.S presentation is to report on the status of that work. However, before reporting on the progress o~ t~at effo~t, 80.8 historical perspective on HVI and prl.Cl.ng studl.es will help to place the current work in the proper context.

Historical Overview

HVI' s evolution has a starting point which is difficult to identify. The Agricultural Research Service, USDA, and other research institutions have been engaged in research and development on instruments for fiber evaluation for many years. One notable point in the evolution occurred in 1976, when the first field trial of instruments was begun in the Lubbock Classing office; this was an experiment undertaken at the direction of the secretary of Agriculture to determine if instruDents could function under the pressures of a classing office environment. The trial was a cooperative endeavor by the Agricultural Marketing Service (AMS), the Agricultural Research Service (ARS), and the Economic Research Service (ERS) of the U. S. Department of AgriCUlture. Since then, the use of instruments first expanded. to Lamesa, where the first fully instrument equipped classing office was established in 1981, then to various other locations , predominantly in Texas and Oklahoma. (It should be noted that instruments have been used in all regions of the cotton Belt for more years than covered here, but not in AMS Marketing Service offices.) During the last two years--since the Cotton Marketing Advisory Committee ' s recommendation to adopt HVI industry wide--AMS has been equipping offices with instrument lines as rapidly as possible.

There has been interest and research on the relations h i p between cotton quality on prices for many years. Several studies between 1916 and 1939 focused on the effect of grades on prices (see references 3, 16, 17, 21, and 24). Later studies addressed the same issue, but with more sophisticated methods (see 15, 18, and 23). Attention on the impact of HVI on cotton prices and the value of fiber properties has grown with the interest in HVI as a grading system, but with some time lag. The initial effort was to determine if instrument grading had a discernable impact on prices, premiums, and discounts (see la, 11, 12, and 22) . These inquiries led to in-depth research in a field described as "hedonic price analysis," a type of analysis which determines the market values of attributes which are embodied in goods. Experience in hedonic analysis in diverse situations with different data (see 1, 2, 4, 5, 6, 7, 8, 9, 13, 14, 19 and 20) led to a realization that the research procedure would be applicable to the AMS fUnction of establishing and reporting prices, premi ums , and discounts.

These ideas were presented to the Cotton Marketing Advisory Committee i n the Summer of 1988 and the Committee endorsed the development and testing of such a system. Since the fall of 1988 , researchers at Texas Tech University have been working jointly with the Market News Branch of the Cotton Division, }.MS, USDA to acconplish that task.

Progress to Date

The primary Objectives are t o design and evaluate an analytical system which AMS can use to support its price reporting activities under the HVI system. The

38

current procedures for gathering and processing market data cannot accommodate HVI grading because of the substantial expansion in the number of quality attributes. The analytical system deve loped requires a computer, the appropriate software, ,and a su!fici~nt volume of mixed lot sales data to perml.t the dell.neatlon of premiums and discounts on the fiber attributes.

Much of the first year was spent on (a) the development a.nd testing of mathematical models which could describe price behavior in a manner which is consistent .... ith existing knowledge about prices, (b) identifying the data requirements, (c) ar r anging for data to run the syste., (d) testing and evaluating the mathematical models, and (e) identifying computer hardware and software needs. Testing of approximately 15 mathematical models has resulted in the selection of two for more extensive evaluation . These formulations are being run on a daily basis by AHS and by Texas Tech. The independent results and evaluations are discussed frequently by telephone. The system requires actual data on individual lot sales, including average lot values for color, trash content, fiber length, fiber strength, etc. (all of the HVI-measured characteristics), price for the lot of cotton, market area (e.g., West Texas ), number of bales in the lot, and date of the transaction. To date, the system has been used only for the Texas- Ok lahoma market because that is the only market in which HVI is already in predominant use and market sales data are readily available (through the cooperation of Plains Cotton Cooperative Association and data from the Telcot system) .

It is anticipated that each of the measured characteristics--trash content, yellowness, grayness, fiber length, strength, length uniformi ty, and micronaire (eventually separated into fineness and maturity)--will have a published premium/discount schedule, although the format for reporting them has not been determined. Alternative formats are being discussed. The information will likely be disseminated through means similar to the current system (the Daily Spot Cotton Quotations), although possibilities for electronic dissemination also exist. All of the premiums and discounts will eventually be incorporated into the Commodity eredi t Corporation loan schedule.

The two models undergoing intensive evaluation are substantially different in their configuration, and each offers some advantages. It is too early to conclude which is the better option, but a decision should be possible by late spring. The computer hardware and software appear to be working well and the people in AHS Market News have demonstrated a command of the models, procedures, and interpretations. Supplementary software which will aid in interpreting and reporting prices, premiums, and discounts is now being developed at Texas Tech.

Several obserVations about the behavior of prices and the system being developed may be relevant to this group. The system generally works--it will decipher market prices and quality premiums and discounts for quality attributes on a daily basis. However, the system requires a larger volume of sales data, and in a more specific form, than does the current system of reporting premiums and discounts on grade, staple, and micronaire. For the two market price reporting regions of Texas and Oklahoma, for example, volume of less than 30-50 lot s8l~s on a given day reduces the reliability of the estlmates. The system must utilize actual data on actual sales--the sales data used cannot be someone 's subjective assessment and they cannot be "partial" data on lot sales.

Analysis to date shows that premiums and discounts are ~ui te variable from day to day. Prices for base quallty seem to be more stable than prices for other qual~ties. Furthermore, the str ength and uniformity preml.ums generally show more day-to-day variability than do the premiums/ discounts for other quality attributes. This situation may not be permanent however, because they could stabilize a fte r quotations ~n the premiums for those attributes are initiated by USDA. The availability of that information should reduce the price volatility becaus~ buyers and sellers are engaging in less "guesslng."

Based on eight months of estimatlng prices on a daily basis, market participants should know that the average premiums and discounts for all attributes in a given market do show daily variation, i.e., they may not be as stable as the spot cotton quotations have implied .

Page 4: Ethridge, D. Development of A Price Reporting System For ...

Unfinished Wgrk

Tasks yet to be accomplished prior to actual implementation include (a) completion of testing the two price models and comparison of their accuracy, (b) selection of the system to be implemented, (C) developing procedures and practices for obtaining and transferring data, and (d) adapting the system to other areas of the cotton belt. Completion of the first three tasks for the Texas-Oklahoma markets appears to be feasible this year. The fourth task is more problematic.

For the system to be applicable for daily price reporting, HVI must be in place in the regions for which prices are being quoted. Consequently, this type of system cannot be implemented in, e. g., the Midsouth, until HVI grading is the norm and HVI data are available in the marketplace . Additionally, the particular mathematical model that works best for the Texas-Oklahoma market may not be the best for markets in other regions. Models for the other regions should be quite similar, but it is reasonable to expect that some modifications may be needed. These matters will have to be tested when the time comes. It will also be necessary to obtain sufficient numbers of sales observations in regions where data are not available from a centralized electronic marketing system . Arrangements must be made with merchants, mills, marketing associations, etc., to obtain the needed sales data on a daily basis.

Acknowledgements

I want to express my appreciation to my co-workers at Texas Tech, especially Ken Bowman, Carlos Engels, Bill Freeman, Melanie Gillis, and Jennifer Stratton , who have contributed and continue to contribute to the progress of the work on a daily basis, and to our co-workers in AHS Market News--Hollis Bowling, Max Parnell, and especially Ron Cole--whose work, diligence, skepticism, and insight has made the work a valuable learning experience. I also want to thank Dale Shaw and others at Plains Cotton Cooperative Association for data assistance and Wendell Barrick, Gary Condra , and Kary Mathis for their suggestions on this paper and its presentation. Financial support for the project is provided by the Market Development Program, AHS, USDA through the Texas Department of Agriculture. Texas Tech University College of Agricultural Sciences Publication No. T-1-318.

References

1. Bowman, K.R. "A Multi-Stage Hedonic Market Model of Cotton Characteristics with Separable supply and Demand." Unpublished Ph. D. dissertation, Texas Tech University, May 1989.

2. Bowman, K.R., and D.E. Ethr~dge. "Pric~s P,~id by Textile Manufacturers for Cotton F1.ber Propert1.es. llll Beltwide COttOD Research Conferences PrQcee~jngs, Cotton Economics and Marketing Conference , Nat1.onal Cotton Council, Memphis, TN, 446-448.

3. Crawford, G.L., and L.P . Gabbard . Prices to Quality of Cotton." Tex. Bulletin No. 383, July 1928.

"Relation of Farlll Agr. EXp. Sta .

4. Ethridge, O. E. "Market Reporting Based on Fiber Properties." 1987 Beltwide cottgn Production. Conference proceedings , National Cotton council, Memph1.s, TN, 44-46.

5. Ethridge, D.E., and B. Davis. "Hedonic Price Estimation for Commodities : An Application to Cotton." IW!le~s~t~e~rrDnU:J~Q~u~r~nn.1ll)-<Q~f'-'A~g~rtl!~c~uU).lt~u~r~ ..... )..J:E~c~QHn~Q~m~~':~c"s 7(1982): 156-163.

6. Ethridge D.E., and K.H. Mathews. "A comparison of Pr1.ce Quotations and Market Prices in the West Texas Cotton Market. " Paper presented at the Western Agricultural Economics Assn. Meetings, July 8-10, 1988, San Diego, CA.

7. Ethridge, D.E., and K.H. Mathews. "Rel~ability of Spot cotton Quotations for Price Discovery 1.n the West Texas cotton Market." Texas Tech Univ. College of Agr. Sciences Pub. No. T-I-212, Aug. 1983.

39

8. Ethridge, D.E., and J.T. Neeper. "Producer Returns From Cotton Strength and Uniformity: An Hedonic Price Approach." SOuthern JournAl of Agricultural Ecgnomics 19(1): 91- 97 (July 1987).

9. Ethridge, O.E., J.T. Neeper, "Producer Prices Received for Cotton Uniformity in the Southwestern U.S ." the Second National Cotton Textile Beach, SC, Nov. 12-13, 1986.

and J. F . Hembree. Fiber Strength and paper presented at

Conference, Myrtle

10. Ethridge, D.E., and D.L . Shaw. "Evaluating the Impact of Instrument Test Line Values on Cotton Marketing and Use." 1977 Beltwide Cotton Research Conferences Proceedings, Cotton Economics and Marketing Conference, National cotton Council, Memphis, TN, 209-212.

11. Ethridge, D.E., and D. L. Shaw. "Results of the Market Impact Evaluation of Instrument Testing at Cotton, 1976 Crop." 1978 Beltwide Cotton Research Conferences proceedings, Cotton Economics and Marketing Conference, National Cotton Council, Memphis, TN, 190-194~

12. Ethridge, D. E., D.L. Shaw, and J.E. Ross. "An Evaluation of the Impact ot Instrument Test Line Values on Cotton Marketing and Use: progress Report ." CEO Working Paper, Economic Research Service, USDA, Aug. 1977.

13. Ethridge, D.E., R. Sudderth, and M. Moore. "A Hodel for Cotton VAriety Selection with yield And Quality Considerations." 1983 8eltyide Cotton Reseprch Conferences Proceedings, Cotton Economics and Marketing Conference, National Cotton Council, Memphis, TN, 293-295.

14. Hembree, J.F., D.E. Ethridge, and J.T. Neeper. "Market Values of Fiber Properties at Southeastern Mill Points." Textile Research Journal 56(1986): 140-144.

15. Horak, J.D . "Cotton Pricing According to Fiber properties ." Natural Fibers Economic Research, Univ. of Texas, Research Rept . No. 110, Dec. 1976.

16. Howell, L.D., and J . S. Burgess. "Farm Prices of cotton as Related to its Grade and staple Length in the united states, Seasons 1928-29 to 1932-33." USDA Tech. Bul. No. 493, Jan. 1936.

17. Howell, L.D., and L.J. Watson. "Cotton Prices in Relation to Cotton Classification Service and to Quality Improvement." USDA Tech. Bul. No. 699, Oct. 1939.

18. Hudson, J.F., and D.R. Williams. "Relationship of Fiber Test Data and Other Factors to Prices Paid for Louisiana cotton." Louisiana state Univ. D.A.E. Research Rept. No. 482, Apr. 1975.

19. Jones-Russell, E., T.L. Sporleder, and H. Talpaz. "Fiber Characteristics Values by Major Spinning Technology." 1286 Qeltwide Cotton Research Conference' Proceedings, Cotton Economics and Marketing Conference, National Cotton COUDcil, Memphis, TN, 337-340.

20. Neeper, J.T. "Producer Values for Fiber Strength and Length Uniformity in the Southwestern Cotton Market." Unpublished H.S. thesis, Texas Tech Univ., Dec. 1985.

21. Paulson, W.E., and J.F. Hembree. "price-~ality Relationships in Farmers' Cotton Markets of Texas . Tex. Agr. Exp. Sta. Bul. No. 501, Dec . 1934.

22 . Robinson, J.A . , D.E. Ethridge, D.L. Shaw, and C.D . Rogers "Market Impact Evaluation ot the Instrument Testing. of Cotton: 1977 crop Progress Report." CEO working Paper, Economics, statistics, and Cooperatives service, USDA, Hay 1979.

23. WilliaMS D. R. "Relationship of Fiber Test Data and Other Factors' to Prices Paid for cotton." unpublished Ph.D. dissertation, Louisiana State Univ., Dec . 1974.

24. and 2 4 ,

Taylor, F. "Relation Quallties ot cotton." 1916.

Between primAry Market Prices USDA Bulletin No. 457, Nov.