Meixia WANG, Ning JIA, Rong YU

Shaanxi Post and Telecommunication College, Xianyang 712000, China

Abstract Taking a total of 903 apple futures data from Zhengzhou Futures Exchange as research samples, and through unit root test, cointegration test, error correction, Granger causality test and variance decomposition with the aid of Eviews 8.0, this paper made an empirical analysis on the relationship between Zhengzhou apple futures market and Chinese apple website (pingguo7.cn) spot market price. It concluded that the main continuity of apple futures has an absolute advantage over the reference value of the futures index, and apple market participants can achieve hedging through the futures and spot market.

Key words Apple futures, Apple spot, Hedging, Empirical analysis

1 Introduction

Apple is an important crop in nearly 25 provinces of China and has wide distribution in China. It is also the main driving force for regional economic development and an important source of income for farmers. However, apple planting is affected by many factors, such as drought, hail, frost and pests and diseases, so the planting risks are high. According to the research of Fu Yaqi[1], in order to solve the "risk" of apple planting and provide hedging tools for fruit farmers and merchants, the futures market can improve the risk resistance ability of the apple industry chain. Apple futures trading is not only conducive to promoting the industrial upgrading of the apple industry and increasing the income of fruit farmers, but also can amplify China’s voices on the international apple stage, laying a good foundation for China to become a global apple pricing center. Therefore, the influence of the development of the apple futures market on the Chinese economy is of immeasurable value.

2 Current research situation

Our research team of apple futures found that although apple futures has been favored by investors since its listing, its open interest and trading volume have continued to grow, and it has become the favorite futures product of Zhengzhou Commodity Exchange, but the apple spot market response is not very strong. An in-depth study of the relationship between apple futures and apple spot, and the role of apple futures market in price hedging is not only helpful for deepening the understanding of the relationship between apple futures and spot markets in China, but also is of great significance to the development of China’s agriculture and rural areas.

3 Review of related research literature

Through sorting out the literature, we found that due to the short trading time of apple futures, there are few related studies. However, there are many studies on related agricultural products, mainly in the following three aspects. (i) Studies on the hedging function of foreign agricultural product futures markets. The research of Li Shengjun[2]shows that the fluctuation of the spot price of maize in the United States is obviously stronger than that of the futures price, and the futures market has a theoretical feasibility for the future value preservation of the spot market. (ii) Empirical analysis and research on domestic and foreign agricultural product futures spot. According to the study of Zhang Youwangetal.[3], no matter long-term or short-term, for soybean and sugar, domestic futures price, international spot price and consumer price index all have a significantly positive impact on domestic spot price; the futures market can attract capital to the agricultural market and use the price discovery function to act on the spot market. Ding Cunzhenetal.[4]made an empirical analysis and found that there is a significant price guidance relationship between domestic and international cotton futures spot markets, but there are differences in the guidance relationship in different periods. (iii) Empirical study on the price guidance relationship between the futures market and the spot market using econometric software. This price discovery function of the futures market is also present in other agricultural products. For example, using Eviews 8.0, Xue Yanjieetal.[5]found that there is a long-term cointegration relationship between corn futures spot prices. Compared with the spot market, the futures market is more responsive to external shocks and plays a leading role in price discovery. They concluded that the futures price has a significant price discovery function and guiding effect on the spot price, while the spot market has no guiding effect on the futures market.

4 Empirical analysis of apple futures spot price discovery function

4.1 Selection and processing of sample dataThe data of apple futures were selected from the closing price of each trading day from December 22, 2017 to March 21, 2019 of the Zhengzhou Stock Exchange intercepted by Straight Flush software. Because the trading time of apple futures is shorter than that of other kinds of products, the reference basis for other data of apple futures price is not obvious. In addition, the time span of a single futures contract is limited. In order to obtain continuous time series data, we selected apple futures main continuity AP9999 and apple spot index AP8888 as reference, which overcomes the time discontinuity of a single futures contract. Furthermore, we selected the daily data of the apple price index in the wholesale market of pingguo7.cn as the spot price of apples, denoted as APPA. The reason for the selection is that the apple price index in the producing area ignores the cost of transportation, and the wholesale price index can better reflect the changing trend of market prices. At the same time, in order to match the trading days of apple futures data, we rejected the data after holidays. Since the compilation base of the Apple Wholesale Price Index is 100, most of the data are around 100 or the number of futures data points vary too much, which loses the meaning of data comparison and makes it inconvenient to observe the trend of changes. Therefore, we expanded the original data by 100 times, and used APPA100 to represent the trend of the spot price of apples. We obtained a total of 903 sample data to study the relationship between apple futures price and spot price with the aid of Eviews 8.0 measurement software.

4.2 Trend and statistical analysis of apple futures and spot price indicesWith the aid of Eviews 8.0 statistical software, we analyzed the variable apple futures price apple main continuity index AP9999, apple spot index AP8888 and APPA100, the results are shown in Table 1.

Table 1 Apple futures and spot price indices

It can be seen that the fluctuation of each index variable is relatively large. In order to prevent heteroscedasticity from affecting the analysis results, we logarithmically processed the original data, and obtained lnAP9999, lnAP8888 and lnAPPA100. From this, we obtained the statistical characteristics of the sample data: the standard deviation of lnAPPA100 was 0.103 825, the skewness was 0.474 705; the standard deviation of lnAP9999 was 0.194 063, the skewness was -0.523 529; the standard deviation of lnAP8888 was 0.191 855, and the skewness was -0.662 543. In terms of standard deviation, apple futures had larger fluctuations than the spot market. The futures market was skewed to the left and the spot market was skewed to the right, neither of which satisfies the normal distribution. The apple main continuity price had a good consistency with the changes in apple index, but the consistency of apple spot was poor. The inflection point of the futures price trend generally occurs earlier than the spot price trend. Thus, it can be known that the futures market price changes earlier than the spot market price, and the futures price leads the spot market to a certain extent. Then, investors can establish hedging through the time difference between apple futures and spot futures market prices, to evade the risks caused by price fluctuations, reduce the risks related to apple, and increase the apple investment returns.

4.3 Empirical analysis

4.3.1ADF unit root test. Since the ADF values of lnAP9999, lnAP8888 and lnAPPA100 were all greater than the critical values of 1% and 5% of the confidence level, and thePvalues were all greater than the given 5% confidence level, the null hypothesis was accepted, and the logarithm of the three groups of data was not stationary. In order to obtain stationary data, we performed first-order difference processing with intercept on the lnAP9999, lnAP8888 and lnAPPA100 data, respectively. To prevent the appearance of such spurious regression results, we performed a unit root test on the data, and the results are shown in Table 2. ThePvalue of the three items was 0.000 0. According to the statistical principle of the unit root, the three groups of data were stationary.

Table 2 Unit root test

4.3.2Johansen cointegration test. According to the above ADF test results, lnAP9999, lnAP8888 and lnAPPA100 were not stationary, while the data ofD(lnAP9999),D(lnAP8888) andD(lnAPPA100) were stationary. We performed Johansen cointegration test on the non-stationary data. Before the cointegration test, it is necessary to use the VAR model to determine the optimal order 4, and then perform the Johansen cointegration 3-order lag test. Test results: the trace statistics value was 56.362 77, the critical value of the given 5% confidence level was 29.797 07, and the null hypothesis None was rejected, indicating that there was at least one cointegration relationship; the trace statistics value was 5.633 31, the critical value of the 5% confidence level was 15.494 71, and the null hypothesis was accepted, so there was only one cointegration relationship. Similarly, the results of the largest eigenvalue test also show that there was one and only one long-term stationary cointegration relationship. Through standardizing the cointegration value, we get the cointegration equation:

lnAPPA100=1.791 344 lnAP9999-1.806 022 lnAP8888

(1)

Through the formula (1), we can obtain that for every 1% increase in the futures main continuity, the spot market will increase by 1.79%; while the futures index rises by 1%, the spot market will fall by 1.81%. According to the above analysis, it can be concluded that the variables were non-stationary variables and there was a co-integration relationship. Then, we further established a vector error correction (VEC) model linking short-term fluctuations with long-term equilibrium. The test results were:

CointEq1=lnAPPA100-1.791 344 lnAP9999-1.806 022 lnAP8888-9.356 494

(2)

It can be seen that the expression of the error correction term and the relational expression of the cointegration test were the same except for the constant term. According to the form of the error correction term, we differentiated the original sequence, and obtained the stationary sequence after the difference to further ensure the stability of the VEC model. The estimated coefficient vector expression was:

(3)

It can be seen that the estimated value of CointEq1 after the error correction in the VEC: the first coefficient value was -0.769, indicating that when AP9999 and AP8888 remain unchanged, the change of APPA100 in periodtcan eliminate 76.9% of the non-equilibrium error in the previous period; the second and third coefficients were close to 0, indicating that AP9999 and AP8888 have little effect on the next period under the change of periodtwhen APPA100 remains unchanged.

4.3.3Granger causality test. According to the test, lnAP9999 was not a Granger cause affecting lnAPPA100.Fvalue was 9.361 80,Pvalue was 5×10-7, rejecting the main continuity of apples in the futures market and the hypothesis that the apple index does not affect the apple spot market, showing that the apple futures market affects the trend of the apple spot market, and the main continuity of apples affects the changes of the apple index; accepting the hypothesis that the apple spot price index does not affect the main continuity of apples and the apple index indicates that the spot market does not influence the trend of the apple futures market.

4.3.4Variance decomposition. (i) Analysis of the impact of the futures market on the spot market. From the first period onwards, the futures market had no impact on the spot market. In the following periods, the fourth period of apple futures main continuity had the most significant impact, reaching 10.05%. The subsequent period started to fluctuate and rose from the fifth period, while the impact of the apple index on the spot market was relatively stable. (ii) Analysis of the impact of the spot market on the futures market. The maximum impact of apple spot market on apple main continuity and apple index was 1.28% and 0.78%, respectively, indicating that the spot had little impact on the futures market. (iii) Analysis of interactions of futures market. The futures main continuity had a great impact on the futures index, the maximum period was 2, and the influence degree was up to 69.46%; the impact of the futures index on the main continuity of the futures was relatively small, and it has been increasing slightly, with a maximum of only 5.56% in the 10thperiod; the impact of futures main continuity on the futures index was greater than the impact of the futures index itself.

5 Conclusions and policy recommendations

5.1 Conclusions(i) In the futures market and spot market, there is a significantly positive correlation between the futures main continuity and the spot, while there is a negative correlation with the futures index. Through the cointegration test, it can be known that the vector error correction model VEC test further proves that the cointegration relationship is valid and that lnAPPA100 is constrained by the long-term equilibrium of other variables in the short term. Besides, from the Granger causality test, we can know that the futures market leads the spot price, and the price discovery effect of futures is significant. (ii) From the differential decomposition, we further obtained that the fluctuation of the futures price has a significant impact on the fluctuation of the spot price, while the spot price fluctuations have little impact on futures fluctuations, and only have a significant impact on their own fluctuations. (iii) The main continuity of China’s apple futures is the most significant for the spot market price discovery in the lag of 4 periods. Compared with the reference value of the futures index, the main continuity of apple futures has an absolute advantage. Apple market participants can achieve hedging through the futures and spot market.

5.2 Policy recommendationsIn theory, it is further proved that the apple futures market does have a hedging function. However, the apple futures market is in the initial stage of development in China. To accept and participate in rational futures investment, market participants need training and education, popularization of basic knowledge of futures, especially hedging knowledge, so that the futures market plays a more effective role. In addition, as a fresh fruit trade product, the taste and texture of apples are not uniform in different planting regions. Therefore, further improving the delivery standards of apple futures is the key to promoting the development of the apple futures market. Also, apple futures need to be combined with other financial products, such as insurance, to promote and influence each other, and further improve the effectiveness of getting rich through apples.