Loading...
Theses
Browse
1 results
Search Results
ThesisItem Open Access Time series modeling and forecasting of tea prices in India(Department of Agricultural Statistics, College of Agriculture , Vellanikkara, 2021) Deenamol Joy; KAU; Laly John, CThe study entitled “Time series modeling and forecasting of tea price in India” was conducted to study the components of time series data on prices of tea in India, to develop time series forecast models for the prices, to develop statistical models for price volatility and to study the integration between international and Indian tea prices. Monthly auction prices of tea for North India, South India and All India for the period from January 1980 to December 2020 collected from the Tea Board formed the main database for the present study. International price of tea for Colombo (Sri Lanka) and Mombasa (Kenya) for the period from January 1980 to December 2020 were collected. To have an idea about the trend in A- Pr- Pd of tea in India, annual data on A- Pr- Pd of tea from 1970 to 2019 in North India, South India and All India were also collected. To have a general idea about trend in A- Pr- Pd of tea in North India, South India and All India, models like exponential, quadratic, cubic etc were fitted. From among several models tried, quadratic model was found to be the best fit for area under tea in North India, while, cubic model was found to be the appropriate fit for production and productivity of tea in North India and, A- Pr- Pd of tea in South India as well as All India. North India and South India tea price data was decomposed to time series components like trend, seasonal variation, cyclic variation and irregular variation. North India and South India showed an overall increasing trend and a prominent seasonal variation. Cyclic variations showed that South India exhibited more cycle of price volatility compared to North India. All India tea price was found to be the simple average of North India and South India tea prices. Compound Annual Growth Rate (CAGR) was estimated for A- Pr- Pd of tea in North India and South India for the period from 1970 to 2019. For North India, growth rate in production was more during 1996-2019 compared to period 1970-1995. For South India, a decline in production was observed during 1970 to 1995. Price forecast models like exponential Smoothing models and ARIMA models were fitted to forecast the tea prices in North India and South India from January 2021 to April 2021. For North India tea price, SARIMA (0,1,3)(0,1,1)12 was identified as the best forecast model whereas for tea price of South India SARIMA (0,1,1)(1,0,1)12 was selected to forecast tea prices. For tea prices in North India and South India, volatility in prices were estimated using intra and inter annual volatility and its significance was tested by fitting suitable ARCH model. Intra annual volatility indices of monthly tea prices in both regions were varying irregularly. In most of the years, North India showed large variation in tea price compared to South India. ARCH (1) model was fitted to check the significance of tea prices and the estimate of ARCH parameter showed high volatility for tea prices for North India and South India. Cointegration analysis was carried out for tea prices to study the integration between international and domestic Indian tea markets. One cointegrating relationship exists between the market pairs, North India - South India, North India – Mombasa and South India – Colombo. No cointegration exist between the market pairs, All India -Mombasa and All India - Colombo. Unidirectional causality was observed between South India and Colombo whereas, bidirectional causality was observed between market pairs, North India - Mombasa and North India - South India.
