Forecasting Farmer Exchange Rate in Central Java Province Using Vector Integrated Moving Average

Abstract

Farmer Exchange Rate (FER) is an indicator that can be used to measure the level of farmers welfare. For every agriculture sector, FER is affected by the historical price of harvest from the corresponding sector and historical prices of other agriculture sectors. In Central Java Province, rice & palawija, horticulture, and fisheries are the largest agriculture sectors which is the main livelihood for most of the population. FER forecasting is a crucial thing to determine the level of farmers welfare in the future. One method that can be used to predict the value of a variable that is influenced by the historical value of several variables is Vector Time Series. An empirical study was conducted using FER data from the rice & palawija, horticulture and fisheries sectors for January 2011-June 2017 in Central Java Province. The results obtained show that by using the VIMA(2.1) model, the FER prediction was very accurate, with MAPE values were 1.91% (rice & palawija sector), 2.44% (horticulture sector), and 2.18% (fisheries sector).

Fuzzy Vector Autoregression for Forecasting Farmer Exchange Rate in Central Java Province

“…The model that we introduce gives MAPE in the range of 5%, so this fuzzy VAR model can be considered for valid use to forecast various time series. Trimono (2020) has made predictions in the same agricultural sector using a vector integrated moving average (VIMA). Although the amount of data used in the VAR study byTrimono (2020) is not the same as the data in this paper, the comparison of prediction accuracy can provide a general view.…”

“…One indicator used to measure the level of farmers'welfare is the Farmer's Exchange Rate (FER) (Trimono et al, 2020). Conceptually, FER appraises the pricing for goods (products) produced by farmers and the cost of goods or services needed to meet their own needs and for their agricultural production process.…”

“…Predictions can be made based on the availability of a number of historical data, especially strong forecasting techniques inferring a stochastic dependence between past and future values (Sagheer and Kotb, 2019). Based on correlation analysis (Trimono et al, 2020) the FER value of an agricultural sector in Central Java Province tends to be correlated with the FER of other agricultural sectors. This is because the harvest price for each sector is linked to the others.…”

“…This paper proposes the idea of a new multiple input and multiple output fuzzy model, namely VAR fuzzy inference. As an empirical comparison of forecast results, the model will be applied to data on the agricultural sector of rice & secondary crops, horticulture, and fisheries in Central Java Province, as has been studied by Trimono et al, (2020).…”

Forecasting Stock Prices on the Lq45 Index Using the Varimax Method

“…Then, Pratama and Saputro (2018) have analyzed and proven the general equation of the VARIMA multivariate time series method by adding exogenous variables to the model so that the multivariate time series method changes to a Vector Autoregressive Moving Average with Exogenous Variable (VARIMAX). Next, Mauludiyanto, et al (2009), who has tested VARIMA modeling with the effect of outlier detection on rainfall data in Surabaya, produced the best model, namely VARIMA (7,1,0) and the relationship between rainfall variables at locations A, B, and C. Trimono, et al (2020) have tested farmer exchange rate in central java province using Vector Integrated Moving Average, which results obtained show that by using the VIMA(2.1) model, the FER prediction was very accurate, with MAPE values were 1.91% (rice & palawija sector), 2.44% (horticulture sector), and 2.18% (fisheries sector). Therefore, from various previous studies and background problems that have been described, this research will discuss forecasting stock prices on the LQ45 index (Liquid 45), namely shares of PT.…”