확장형 ARMA-GARCH 모형을 활용한 농업관측사업의 가격안정화 효과 분석

Abstract

The agricultural outlook organized by KREI Agricultural Outlook Center has been implemented since 1991 as one of the representative policies aimed at stabilizing agricultural supply and demand. And the scale of Agricultural Outlook is steadily expanding. The price stabilization effect analysis, which is noted in this study, has been steadily measured in previous studies as one of the methods of analyzing the effects of Agricultural Outlook. In the preceding studies, the effect of price stabilization was analyzed by various methodologies, but consistent results were obtained in that the price fluctuation rate was lowered after the Agricultural Outlook. However, price drops occurred periodically for five sensitive items, including onion, garlic, pepper, cabbage, and radish. Therefore, it is difficult for farmers to easily agree on the price stabilization effect. In recent years, prices have plunged due to overproduction of onion and garlic, which is one of the biggest issues in agriculture. The Agricultural Outlook has always been mentioned as a cause and alternative to this situation. The drop in onion and garlic prices is similar. In other words, the public demand for sophisticated Agricultural Outlook is increasing, and the budget to be invested in Agricultural Outlook in the agricultural budget for next year was announced to be about 17 billion won. This is twice the size of the 8.5 billion won budget, which was last year. Therefore, next year's Agricultural Outlook is expected to take a new turnaround, and it is time to discuss how to improve agricultural outlook. For this discussion, it is necessary to examine the trend of price stabilization effect according to the advancement that has been made. Based on the measured analysis results, it is possible to judge the effectiveness of the advancement that has been carried out so far, and can be used as fundamental data for the future development direction of Agricultural Outlook. Based on this necessity, this study analyzed the price stabilization effect on five sensitive items. In this study, the timing of the effect of agricultural outlook was detected and viewed as the point of intervention. At this time, the level shift outlier detection, which is one of the methods for detecting outliers, was used. The price stabilization effect was measured based on this time point, and the ARMA-GARCH model, which is a time series model assuming conditional heteroscedasticity, was used. The ARMA model, which is an average equation of price, reflects the effects of seasonality, agricultural market opening, disposal policy, and abnormal climate through dummy variables. The price variance equation measures the trend of the price stabilization effect by multiplying the trend variable with the dummy variable considering the point of intervention of agricultural outlook. The analysis results for level shift outlier detection are as follows. For onion and garlic, there was a change in time-series price fluctuations within 5 years after the agricultural outlook was implemented. On the other hand, there was no such level shift for dried peppers, cabbages and radishes. Dried red pepper farmers are mostly small farmers, and most of the cabbage and radish are pre-traded in the field. Accordingly, it was interpreted as being less affected by the agricultural outlook. Therefore, for onion and garlic, the trend of price stabilization effect was measured by ARMA-GARCH model. Although the level of volatility did not change for dried pepper, cabbage and radish, the trend of price stabilization effect was measured from January 1999 when the Agricultural Outlook was implemented. At this time, the estimation coefficient of the dummy variable reflecting the trend was significantly estimated as a negative value for onion, garlic, cabbage, and radish. This implies that price volatility has also decreased significantly as agricultural outlook is implemented and developed. On the other hand, the dried pepper was not significantly estimated. The above results are summarized as follows. Onion and garlic showed significant changes in price volatility levels within five years after the agricultural outlook, and price volatility declined thereafter. In the case of cabbage and radish, there was no significant level shift in the price fluctuation rate, but the price volatility decreased gradually after the agricultural outlook. In case of dried pepper, there was no significant change in price volatility level, nor did price volatility decrease significantly. These results are similar to those divided by the share of forward contract. Looking at the proportion of these transactions, onions and garlic are 30-50%, cabbage and radish are 70-90%, and dried pepper is 0%. Further analysis can be made on the correlation between the portion of forward contract and price stabilization effects. The significant results of this analysis are expected to be useful resource to upgrade Agricultural Outlook projects. Therefore, it is necessary to consider differentiated strategies by item when discussing future development plans. In particular, distribution methods, farm size, and cropping systems should be considered for the five sensitive items, which are subject to severe price fluctuations and plunges periodically. For example, dried peppers, unlike other items, are not pre-traded in the field and are mostly small farmers. Accordingly, the results of this study also showed no change in the level of time-variable price fluctuations in dried pepper, and even after the implementation of the agricultural outlook, price volatility did not tend to decrease. As the project budget for agricultural outlook will double in next year, Department of agricultural outlook in KREI should consider the development direction in this respect. On the other hand, this study did not completely exclude the possibility of overlapping with other agricultural policies in detecting the level shift outliers that the effect of the Agricultural Outlook has. In addition, this study focused on the variance equation of price using the ARMA-GARCH model, so that the trend change is known, but there is a limit in accurately measuring the magnitude of the change in price fluctuation rate. However, this study is based on a new approach to the point of intervention for Agricultural Outlook that has been uniformly assumed for all items, and the trend is measured in consideration of monthly, agricultural market opening, disposal policy, and abnormal climate in the ARMA-GARCH model.

농업관측사업이란 안정적인 농산물 수급을 위해 재배면적, 작황, 생산, 출하, 재고, 수출입 등 국내외 정보를 종합적으로 관측하여 관련 종사자에게 제공하는 사업이다. 하지만 채소 품목의 경우 주기적으로 가격 폭락이 발생하고 있으며, 특히 작년부터 불거진 양파·마늘 가격 폭락 사태는 지금까지 이어지고 있다. 이에 따라 정밀하고 시의성있는 농업관측 고도화에 대한 국민적 요구는 높아지고 있다. 이를 반영하듯 내년도 농업관측 예산이 85억 원에서 169억 원으로 2배 가량 늘어났으며, 농업관측사업이 새로운 국면을 맞이할 것으로 기대된다. 따라서 본 연구에서는 그간 이뤄졌던 농업관측의 발전에 따라 가격안정화 효과가 어떤 양상으로 나타나고 있는지 살펴보고, 추후 농업관측 고도화를 논의할 때 시사점을 제시하고자 한다. 가격안정화 효과 분석은 농업관측의 사업효과를 계측하는 대표적인 방법 중 하나로서 선행연구에서도 꾸준히 계측되고 있다. 하지만 분석 품목과 분석 모형에 있어 조금씩 차이가 있으며, 관측사업 사업효과가 나타나는 시점에 대해서는 일률적인 가정을 하고 가격안정화 효과를 계측하였다. 이에 본 연구에서는 관측사업의 개입시점을 고려하여 분석을 시도했으며, 가격안정화 효과 분석에는 ARMA-GARCH 모형을 이용하였다. 이때 농산물의 계절성, 이상기후, 산지폐기, 수입개방 등의 영향을 더미변수를 이용하여 고려하였다. 분석 품목은 채소 5대 민감품목인 양파, 마늘, 건고추, 배추, 무로 한정하였다. 분석 결과, 양파와 마늘은 농업관측사업이 시행되고 5년 내 가격변동률 수준에서 유의한 변화가 나타났으며, 이후에도 가격 변동성이 감소하는 추세를 보였다. 배추와 무는 관측사업 이후 가격변동률 수준에서 유의한 변화가 나타나지 않았지만, 가격변동성이 유의하게 감소하는 추세를 보였다. 건고추는 가격변동률 수준에서 유의한 변화가 나타나지 않았을 뿐만 아니라 가격 변동성이 감소하는 추세도 보이지 않았다. 추후 발전방안을 논의할 때, 이러한 결과를 토대로 품목별로 차별화된 전략을 고민해볼 수 있다. 예를 들어 포전거래율과 가격안정화 효과 간 상관관계가 나타났기 때문에 이러한 부분을 고려하여 농가보다 산지유통인에 초점을 맞춰 정보를 전달하는 등 여러 가지 발전 방안을 모색할 수 있다.