Changes in Crop Prices before and after Releases of Statistics Canada's Field Crop Reporting Series Publications^{Note *}

View the most recent version.

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

Archived

This page has been archived on the Web.

by

Bishnu Saha
Agriculture Division, Statistics Canada
Jean Talon Building, 12th floor
170 Tunney's Pasture Driveway
Ottawa, Ontario K1A 0T6

and

M. Ashraful Alam
Department of Environment and Geography
University of Manitoba
Winnipeg, Manitoba

December 2011

Skip to text

Abstract
Introduction
Objectives
Data and methods
Results and discussion
Conclusion

Text begins

Abstract

Statistics Canada administers six surveys per year to collect information on intended, seeded and harvested acreages, yields, production and stocks of principal field crops, and publishes these survey estimates in the Field Crop Reporting Series (FCRS). This paper analyses short-term movements in weekly crop prices from the week before the releases of FCRS to the week after the releases. Field crops included in this study are oats, canola, corn, flax, barley and wheat, while specialty crops studied are sunflower seed, canary seed, field peas, lentils, mustard seed, chick peas and green peas. The data for field crops cover a period from 1990 to 2009 and that for specialty crops cover varying periods from 1992 to 2009 based on their availability. The results reveal that the price changes before and after the official releases of FCRS tend to even out over time. The results also suggest that prices after the releases are as likely to increase as they are to decrease. Based on the findings, the study concludes that the publication of statistics in the FCRS has no systematic effect on crop prices. The results are consistent with the findings of the National Agricultural Statistics Service of the United States Department of Agriculture.

Introduction

Statistics Canada administers six surveys per year to collect information on intended, seeded and harvested acreages, yields, production and stocks of principal field crops, and publishes these survey estimates in eight publications of the Field Crop Reporting Series (FCRS). Possible market reactions to these publications are sometimes a source of controversy among farm operators. In theory, improved market information is likely to increase markets efficiency by reducing transaction costs, thereby raising prices farmers receive for their crops (Hoffman, 1980). However, many farm operators are not convinced that the releases of crop reporting series produce the desirable market results. They often voice concerns that crop markets do not react efficiently to the information and argue that crop prices fall more often than they rise after publication of the estimates. Sometimes their concerns lead to complaints to the effect that the published estimates generally bring prices down.

In agricultural commodity markets, new information induces traders to re-evaluate their current expectations about future market conditions involving a variety of factors. These factors include, among others, global inventory of a commodity, global production, global supply and a country's contribution to the global supply. Identifying specific sources of price volatility in agricultural commodity markets is an important component in developing a marketing strategy for agribusiness firms. The publication of new information on the supply side can be one of many factors in price adjustments if the information is unanticipated or contrary to traders' previous expectations (Fama, 1970).

In an efficient market, price changes should reflect differences between market expectation of new information and the actual information upon release (Colling and Irwin, 1990). As market fundamentals are expected to be instrumental in determining prices, market participants are likely to respond to new information on supply and demand conditions. According to efficient market hypothesis, prices in an efficient market ought to reflect all known information available at a given time (Fama, 1970). When prices do not change or change slightly after the release of new information, there are reasons to believe that either the markets have accurately anticipated the information or the impact of the new information on the market is infinitesimal.

Canadian grain, oilseed and specialty crop prices react to a multitude of influences. These influences originate from a wide variety of sources and fall into general categories such as environmental, economic and political (Wilson, 2003). Most grains, oilseeds and specialty crops in Canada are grown in the Prairie provinces of Alberta, Saskatchewan, and Manitoba. Almost all of these grains and oilseeds are grown under dryland conditions with a very short growing season. Yields vary significantly with temperature and precipitation at critical stages of crop development, and therefore, production and domestic supply often remain uncertain. On the demand side, prices react to variations in demand owing to economic and demographic factors.

In recent decades, globalization, market pressures, technology and innovations have spurred Canadian agriculture to increase output and productivity. On the other hand, increased global production by efficient means has contributed to a long-term decline in real commodity prices. Various forms of subsidies especially in countries with larger shares of global production also place downward pressure on world prices. As Canada is an export-oriented nation and a price-taker in the international commodity markets, these global trends have a significant impact on Canadian prices.

Objectives

Given the concerns often voiced by farm operators as survey respondents, the Agriculture Division of Statistics Canada has an interest in understanding the effect of its crop reports releases on Canadian crop prices. The primary objective of this study is to analyse short-term movements in weekly crop prices from the week before the releases of Statistics Canada's Field Crop Reporting Series to the week after the releases. The results will provide an opportunity to compare the statistics of crop price changes before the release dates with those after the release dates. The results will also provide a basis for further in-depth analysis of price movement patterns, if there is any.

Data and methods

In this study, the effects of crop reporting series on prices of major grains, oilseeds and specialty crops are investigated. Specifically, price movements in the futures markets immediately before and after the release of crop reports are studied. The major grains studied are oats, barley and wheat, while oilseeds include canola and flaxseed. The specialty crops included in this study are large green lentils, small green lentils, medium green lentils, French green lentils, red lentils, brown mustard, yellow mustard, oriental mustard, sunflower, canary seed, desi chick peas, Kabuli chick peas, field green peas, field yellow peas and field peas.

Data on weekly commodity prices are obtained from Saskatchewan Agriculture and Food. The data period extends from 1990 through 2009 for major grain crops except oats (1993-2009) covering a total of 750 releases of the Field Crop Reporting Series. The data period for specialty crops is from 1992 to 2009, except for three specialty crops including red lentils, desi chick peas and Kabuli chick peas. The data period for these three specialty crops ranges from 1999 to 2009. For the specialty crops, the study period covers a total of 1,856 releases of the Field Crop Reporting Series.

Weekly prices of each crop on the week immediately before release, on the week of release and on the week immediately after release are tabulated. The changes in prices from the week before to the week of release (i.e., changes before the releases) and from the week of release to the week after (i.e., changes after the releases) are recorded as increases, decreases and no change. Price changes for each crop are then grouped by changes before the releases and changes after for comparison.

At the second stage, percentage changes in weekly prices from the week before to the week of release and from the week of release to the week after are analysed. Descriptive statistics of percent price changes are derived and the minimum, the maximum, the mean and the standard deviation of percent price changes before and after the releases are recorded for each crop. The upper and the lower limits of the 95% confidence interval for percentage changes in weekly prices before and after the releases are then estimated.

Finally, a chi-square test is conducted to check the probability that the observed frequencies of increases and decreases are equal. The chi-square test is performed to determine whether the null hypothesis as stated below is supported or rejected. The null hypothesis is that the price reactions are negative and positive in equal proportions of time. Therefore, the alternative hypothesis is that the distribution is not the same as stated in the null hypothesis. The null and alternative hypotheses are as follows:

H₀: The price reactions are negative and positive in equal proportions of time.
H₁: The distribution is not the same as stated in the null hypothesis.

Results and discussion

The empirical evidence suggests that Statistics Canada's Crop Reporting Series releases do not have any significant effect on prices of either specialty crops or major grains and oilseeds (Table 1 and 2). Changes in weekly prices from the week before release to the week after release have been studied for a total of 15 specialty crops. Table 1 illustrates the price changes for all specialty crops categorized as increases, decreases and unchanged. For specialty crops on average, from the week before to the week of release, prices remained unchanged in 52.3% of the times, increased in 25.4% and decreased in 22.3% of the times. Almost a similar pattern is observed between the week of release and the week after the release. Prices remained unchanged in 53.0% of the times after release, increased in 24.2% and decreased in 22.8% of the instances. Figures 1 through 20 depict graphical comparison of weekly price changes before and after releases for each crop studied.

For major grains on average, percentages of increases and decreases between the week before and the week of releases were almost equal (49.5% vs. 48.8%; Table 2). Only in 1.7% cases, prices remained unchanged. Between the week of release and the week after release, the percentage of increases was noticeably higher than decreases. During the time, prices increased 52.7 % and decreased 46.1% of the times.  This lends some support to the fact that prices tend to stick to their upper limits and adjust to a lower level only when the information is unanticipated or contrary to traders' previous expectations.

From Table 1 and 2, it can be concluded that changes in crop prices are as likely to increase as to decrease and price movements are not systematically associated with Statistics Canada's publications in the Field Crop Reporting Series. The results do not show any evidence that Field Crop Reporting Series publications have an exclusively negative bearing on market prices. 

Table 3 and 4 show descriptive statistics for all major and specialty crops. It is apparent from Table 3 and 4 that mean percentage of price changes for both major grains and specialty crops are not significantly different from zero. On average, price changes after the publication of Crop Reporting Series are not significantly different than those before the publication. However, a closer look at the numbers for percentage changes reveals that from the week before to the week of release the mean percentage change was negative for 6 out of 15 specialty crops (Table 3). From the week of release to the week after, the number of negative means was reduced to 3 (Table 4). For major grains, the number of negative means reduced from 2 to 0. These results do not support the claim that prices fall consequent to releases in the Field Crop Reporting Series.

Table 5 and 6 show the observed and expected frequencies to be used in the chi-square distribution test. Observed frequencies are actual increases and decreases in prices, while the expected frequencies refer to the expected values when the increases and decreases are equal. Since there are two occurrences of price changes (increases and decreases) for each crop, the degree of freedom is 1. Thus, the critical value at the 95% confidence level is 3.84. The chi-square values for crops calculated from actual and expected frequencies have always been less than the critical value of 3.84, except the price changes of Field Yellow Peas from the week before to the week of release (Table 5) and Yellow Mustard from the week of release to the week after (Table 6). The chi-square values for the sum of the incidents of price increases and decreases after the Field Crop Reporting Series releases for major grains and specialty crops are also lower than the critical value (Table 5 and 6).

Since the chi-square values in general are lower than the critical value, there is not enough evidence to reject the null hypothesis that the price reactions are negative and positive in equal proportions of time (i.e., price reactions to the official releases of Field Crop Reporting Series publications tend to even out over the time). For specialty crops, however, the chi-square value for the sum of the instances of price increases and decreases before the releases is significant at the 10% level. This indicates that in cases of some specialty crops, price increases due to market factors (and perhaps due to expectations) before the releases. Another observation that supports this result is that for most crops, the ratio of positive and negative price changes before the releases does not vary significantly from that after the releases (Table 3 and 4).

Conclusion

This study evaluates the impact of public information relating to production on market prices of crops by analysing price changes before and after releases of Statistics Canada's Field Crops Reporting Series. As the results indicate, there is no evidence to support the claim that crop releases have a negative bearing on market prices. The results also suggest that mean price changes for both "week before to week of release" and "week of release to week after release" are not significantly different from zero. Furthermore, the interval estimates of the mean percentage of price changes are short and are approximately symmetric about zero. Chi-square tests conducted for all crops also exhibit that prices are as likely to increase as to decrease during the week after the releases of Statistics Canada's Field Crop Reporting Series.

Based on the above findings, it is concluded that official releases of Statistics Canada's Field Crops Reporting Series publications do not have any systematic impact on market prices of major grains, oilseeds and specialty crops. 

---

Note

* The supervisory support of Verna Mitura (erstwhile Unit Head, Research and Development Unit) in setting up the study and the support of the divisional management for its continuation are acknowledged.