© The Authors, published by EDP Sciences, 2021

1 Introduction

The dynamics of meat consumption in Russia differs from trends in other significant markets. The level of meat consumption is based on an estimate of the growth of incomes of the population. In 2020, meat consumption in Russia (in dressed weight) reached a record of 77 kilograms per capita. A meat consumption is stabilizing in Russia and meat production is growing.

The demand for products is determined by the purchasing power of the population and other subjective factors. The regional meat market is a system in which good exchange between producers (sellers) and customers. The research of market conjuncture involves the study of its development in the near future to determine what measures should be taken in order to better satisfy the demand of the population for goods, to more rationally use the opportunity to increase exports while enriching its own market and ensuring the food security of the region. Constant study and forecasting of the market conjuncture is a prerequisite for the successful operation of the region.

2 Methods

The research methods of the regional market were used in the works of such scientists as A. Belousov, V. Oreshin, Y. Vertakova, E. Kuzbozhev.

The meat consumption is stabilizing in Russia and meat production is growing. The conjuncture of the meat market of the Kursk region is characterized by the ratio of supply and demand for meat and meat products, as well as by the level and ratio of prices [3,6,7,8,13,16]. The regional meat market is a system in which good exchange between producers (sellers) and customers. The target benchmark for the development of the meat complex can be the expected volume of demand for the industry’s products. The research of market conjuncture involves the study of its development in the near future to determine what measures should be taken in order to better satisfy the demand of the population for goods, to more rationally use the opportunity to increase exports while enriching its own market and ensuring the food security of the region [14,16,17].

The regional meat market is part of the all-Russian meat market and part of the agricultural and food markets. Its main role is determined by the large volumes of manufacture and consumption of this group, as well as the importance of meat and meat products as the main sources of animal proteins in the human diet [1,2,3]. In the Kursk region, the supply of meat products from commodity producers is four times higher than consumption. The target benchmark for the development of the meat complex can be the expected volume of demand for the industry’s products. Constant study and forecasting of the market conjuncture is a prerequisite for the successful operation of the region [1,3,7,8, 15,16].

In this regard, the conjuncture of the meat market of the Kursk region is characterized by the ratio of supply and demand for meat and meat products, as well as by the level and ratio of prices [3,6,7,8,13,17]. The research of market conjuncture involves the study of its development in the near future to determine what measures should be taken in order to better satisfy the demand of the population for goods, to more rationally use the opportunity to increase exports while enriching its own market and ensuring the food security of the region. [3,6,7,8,17,13]. Constant study and forecasting of the market conjuncture is a prerequisite for the successful operation of the region [1,3,7,8,13,15,16].

The target benchmark for the development of the meat complex can be the expected volume of demand for the industry’s products.

The interrelationship of demand for a product and its determinants is reflected by the general demand function and is presented in the form of a multiple regression model:

$$\text{QDx}\hspace{0.17em}=\hspace{0.17em}\text{f}\hspace{0.17em}\text{(Px,}\hspace{0.17em}\text{Py,}\hspace{0.17em}\text{I,}\hspace{0.17em}\text{Tx,}\hspace{0.17em}\text{Inx,}\hspace{0.17em}\text{Psax,}\hspace{0.17em}\text{q),}$$(1)

where QD_x is the volume of demand for meat and meat products per unit of time;

P_x average product price, in rubles;

P_y, is the price of a substitute product (fish), in rubles;

I the average income of the customer, in rubles. per year;

T_x is the average meat consumption per person, kg per year;

Inx consumer price index, at times;

Pasx structure of consumption expenses on products, in %;

q is the number of consumers of products.

The primary statistical information was brought to a level where it is possible to build a regressive addiction of the volume of demand on the above-mentioned factorsigns using the MathCADProfessional package [9,10].

The forecast оf the demand for meat in the Kursk region was obtained using neural networks (NN), since NN technology is a rather flexible tool for predicting various indicators and can be used to forecast demand [2,3,8,11,12].

The neural network tuning process was carried out with repeated running of the training examples until the error value decreased to acceptable limits. The configured network was used for forecasting using the Neural Network Emulator program.

The neural network was trained under the following conditions and restrictions (neural network configuration):

• sigmoid parameter: 0.5 (used to provide nonlinear data conversion);

• number of entrances: 7;

• number of hidden layers: 1;

• layer 1 number of neurons: 6;

• number of outputs: 1;

• learning speed: 0.1;

• moment: 0.9;

• criteria for stopping learning: 10,000 epochs have passed,

• using the test set as a validation one, that is, training was stoped with the appearance of a message as soon as the error on the test set begins to increase.

3 Results and Discussion

The dynamics of meat consumption in Russia differs from trends in other significant markets. The level of meat consumption is based on an estimate of the growth of incomes of the population. In 2020, the average annual consumption of animal proteins in our country was 25.1 kilograms per capita, which is one kilogram more than in 2019. In 2020, meat consumption in Russia (in dressed weight) reached a record of 77 kilograms per capita. On average, the consumption of meat in the world is 43 kilograms per person per year, and, according to the UN, in rich countries it is 83 kilograms [1].

To construct the supply and demand curves, the following data were required, presented in table 1.

Supply and demand curves are shown in Figures 1 and 2.

The demand curve has been changing since 2015, but sales volumes are constantly growing and do not depend on consumer prices. An increase in demand with an increase in prices is observed in the case of indispensability of products, and this happens in the meat market. In this case, the elasticity of demand is above zero.

The supply curve since 2017 has been “classic”, that is, a positive slope. The supply and demand curves do not intersect.

Let’s Form the elements of a series of retrospective data on the volume of demand obtained by multiplying the population of the Kursk region by the actual consumption of meat and meat products on average per person per year. The retrospective period is 2005 2020 [5, 6, 7, 8].

Factor signs affecting demand will be reduced to a comparable form (Px, Py, I) and are given in 2020 prices.

This procedure requires data of consumer price indices (CPI) (Table 2).

Having retrospective data about indices and cost indicators, we calculate the indicators taking into account the consumer price index (Table 3).

The primary statistical information was brought to a level where it is possible to build a regressive addiction of the volume of demand on the above-mentioned factorsigns using the MathCADProfessional package.

As a result of data processing by methods of regression analysis, the following results were obtained:

• multiple correlation coefficient R = 0.893, determination coefficient R2 = 0.782;

• the significance of the coefficients of the regression equation was checked using the Student’s test (tstatistics). The following t-test values were obtained: b114.76; b2-14.77; b3-10.28; b4-14.56; b5-14.78; b6-14.66; b7-13.56.

We represent the solution of the system of linear equations in matrix form:

$$\begin{array}{l}Y=\left\{\begin{array}{l}\begin{array}{c}251356\end{array}\\ 410400\\ 497250\\ 531424,8\\ 690525\\ 685343,7\\ 731952\\ 810652\end{array}\right\}\text{X}\\ =\left\{\begin{array}{cccccccc}1& 72,5& 45,5& 15106& 47,0& 1,2& 11& 5348\\ 1& 125,3& 89,3& 19566& 51,3& 1,5& 8,5& 8000\\ 1& 199,0& 91,6& 21060& 68,5& 3,2& 9,0& 8500\\ 1& 201,0& 100,7& 25330& 59,0& 4,6& 9,0& 9002\\ 1& 196,8& 124,4& 26111& 75,0& 0,9& 8,2& 9207\\ 1& 225,2& 128,8& 27275& 70,1& 1,9& 8,2& 9207\\ 1& 209,5& 130,2& 29247& 78,0& 4,3& 8,5& 9384\\ 1& 211,3& 135,5& 35225& 82,0& 4,9& 9,4& 9586\end{array}\right\}\end{array}$$

The formula for calculating the vector of regression coefficients in the vector-matrix notation has the form [9, p. 311]:

$$b={(X^T\cdot X)}^{-1}{X}^T\cdot y,$$(2)

We obtain the vectors of estimates of the regression model and, based on the obtained vector of estimates of the regression model, we compose an additive function:

$$y=-7,563\cdot {10}^4+7,593\cdot x_1-10,541\cdot x_2+0,021\cdot x_3=1.336\cdot {10}^3\cdot x_4-20.637\cdot x_5+70.261\cdot x_6+55.799\cdot x_7$$(3)

The next step of the study will be obtaining calculated values for the function and the relative error, which we will summarize in Table 4.

The relative error S is determined by the next formula:

$$s=\frac{y_{actual}-y_{settlement}}{y_{actual}}\cdot 100\%$$(4)

where У_actual is the actual value of demand for products,

У_settlement is calculated value.

The volume of the array of values of the input variables was limited to 5 years (Table 6). The forecast values for the next 2 years (2021, 2022) will be calculated using the Statistica 6.0 package. The real values of the input indicators were entered into the program and the predicted values of the QDx variable were obtained:

Q2021=860321 tons; Q2022=883254 tons.

In the course of training, the following was achieved: the value of the average error of the training sample is 0.0001204, the test sample is 0.0003134. The maximum error of the training and test sample was 0.00108 and 0.00106, respectively.

Fig 1. Demand curve

Fig 2. Supply curve

4 Conclusions

According to the data obtained, we can conclude that the demand for meat and meat products is growing in the Kursk region, the market is developing, and business activity is quite high.

This is a positive trend despite the fact that solvent demand is unstable. An optimistic forecast of growth in demand for meat in the short term has been obtained.

References

All Tables

All Figures

	Fig 1. Demand curve
In the text

	Fig 2. Supply curve
In the text