PhD

This study investigates price behaviour, volatility, and forecasting of coconut, copra, and coconut oil prices in Kerala, while also evaluating marketing efficiency and market structure. Time series analyses were conducted using monthly price data of coconut, copra, and coconut oil from 1980 to 2022. Time series decomposition revealed that prices of all three commodities were typically higher during lean supply months and lower during peak harvesting seasons. Coconut oil in domestic markets showed the steepest upward trend among all markets studied, including international and edible oil prices. Cyclical trends were better defined in the pre-liberalisation period (1980–2002), with cycles lasting 4–5 years, but became shorter and more frequent (2–3 years) during the post-liberalisation period (2003–2022). Structural break analysis using Bai and Perron's method identified 2002 as a pivotal point, which warranted the division of the dataset into two sub-periods for comparative analysis.
Volatility analyses revealed different outcomes. Intra-annual price volatility declined for most coconut products in Kerala during Period II, although exceptions like ball copra and dry coconut in Kozhikode showed increased volatility. Internationally, coconut oil volatility declined, while copra prices showed heightened intra-annual volatility. Edible oils, except palm kernel oil, also experienced rising volatility during this period. Inter-annual volatility for domestic coconut products and international edible oil markets, except palm kernel oil, decreased in the post-liberalisation period, contrary to the international prices of copra and coconut oil, which became more volatile. Coppock’s instability index confirmed these trends, indicating that the post-liberalisation period saw reduced annual instability for most coconut products except copra, although domestic instability remained lower than that in international markets. The Generalised Autoregressive Conditional Heteroscedasticity (GARCH) (1,1) model, used to assess the volatility persistence, produced estimates that were statistically insignificant across all commodities and markets, suggesting a lack of long-term volatility despite observed fluctuations.
Price integration analyses using Johansen’s cointegration and Granger causality tests provided critical insights into market linkages. Within Kerala, strong long-term integration existed among the coconut, copra, and coconut oil markets. Market leadership shifted from Alappuzha and Kozhikode in the pre-liberalisation era to Kochi in the post-liberalisation period. At the interstate level, moderate to strong cointegration was observed between Kerala and key markets in Tamil Nadu and Karnataka, with Arisikere and Kangayem emerging as influential markets in the price formation of copra and coconut oil. Internationally, the pre-liberalisation period exhibited moderate integration of Kerala’s markets with international coconut product prices, but this weakened significantly in the post-liberalisation period. Granger causality analyses revealed unidirectional influences from Kochi to global markets in Period I, with no significant linkages in Period II. No consistent cointegration or causality was found between Kochi coconut oil prices and international edible oil prices or edible oil import prices, implying that trade liberalisation did not significantly improve transmission of price signals from global markets to Kerala’s coconut economy.
The performance of various forecasting models viz., regression, exponential smoothing, Autoregressive Integrated Moving Average (ARIMA), GARCH, Artificial Neural Network (ANN), Vector Auto Regression (VAR), and a hybrid Convolutional Neural Network - Long Short-Term Memory (CNN-LSTM) models was evaluated using monthly price data from 1990 to 2022, with forecasts generated for the year 2023. Forecast accuracy was assessed by comparing predicted prices with actual observed values, using Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). Holt-Winters Multiplicative Seasonal (HWMS) model yielded the best performance for coconut without husk in Alappuzha, coconut oil across multiple markets and milling copra prices in Kozhikode and Alappuzha. ANN model was judged best for the prices of milling copra in Kochi, while the ARIMA model outperformed others in predicting the prices of dry coconut and ball copra prices in Kozhikode.
A micro-level survey was conducted in two blocks, each from Kozhikode, Malappuram, Kannur, and Thrissur, to study marketing practices. The majority of farmers sold raw coconuts (90%), while only 10 percent of farmers were involved in processing. Local traders dominated procurement in most districts, except Thrissur, where copra makers played a significant role. Ten marketing channels were identified. Channel IX, in which farmers processed coconuts into oil and sold directly to consumers, was the most efficient one (5.53). Contrary to this, the trader-dominated channels had lower efficiency, with Channel I (comprising husked coconuts) being the least efficient (2.96). Channels involving cooperatives and government support performed better in terms of marketing efficiency and producer returns. Price fluctuations were the most cited constraint by farmers, discouraging long-term investment in coconut farming. Other challenges included high labour costs, intermediary charges, and limited transportation facilities. Farmers rarely stored or processed coconuts due to the risk of spoilage and a preference for immediate sales.
Market structure was analysed using the Structure-Conduct-Performance (SCP) framework. The Herfindahl-Hirschman Index (HHI) value of 451.53 and CR₄ value of 31.70 percent indicated a competitive structure with minimal concentration for village traders. In Vadakara, a key market for dry coconut and ball copra wholesale market, the HHI value of 873.22 confirmed a competitive structure. However, the CR₄ value of 45.21 percent suggested moderate oligopsonistic behaviour. Wholesalers sorted and processed ball copra into Rajpur copra and sold dry coconuts to North India through commission agents. However, constraints such as dwindling supply, high labour costs, storage issues, and financial instability due to delayed payments remained prevalent. The coconut oil processing and export sectors displayed a moderately concentrated and oligopolistic structure (HHI of 1799.70 and CR₄ value of 83.28%). While small processors handled limited volumes, larger firms sourced coconuts from distant districts. Marketing costs and margins were ₹1,150 and ₹1,510 per quintal, respectively. The exporters achieved margins of ₹6,511 per quintal (28% of the sale value). However, the sector faced stiff competition from Tamil Nadu, an influx of adulterated coconut oil, a shortage of copra and shifting consumer preferences. KERAFED, Kerala’s apex coconut farmers' federation, played a pivotal role in procurement and processing, producing 14,000 MT of coconut oil in 2020-21. Despite increasing procurement volumes, the high operational cost (92–99% of revenue) and minimal value addition limited the profitability of KERAFED. The federation faced multiple constraints, including inactive primary cooperatives, dependence on Tamil Nadu for copra, labour shortages, and inefficiencies in implementing procurement schemes.
Surveys among five Coconut Producer Companies (CPCs) revealed diverse operations. Thrissur CPC focused on Neera and its by-products, while Ponnani CPC specialised in seedlings. The remaining CPCs engaged in oil processing. Neera marketing efficiency was low (1.05) due to high costs and limited demand. Vadakara CPC, the most active in oil processing, achieved a marketing efficiency of 2.82. Despite engaging in value addition (virgin oil, chips, desiccated coconut), CPCs faced challenges like high administrative costs, competition, and discontinuation of government support.
The study concludes with actionable policy recommendations to enhance stability and efficiency in Kerala’s coconut economy. Establishing a coconut market intelligence cell is proposed for monthly price forecasting using HWMS, ARIMA and ANN models. A farmer-friendly mobile application is recommended for disseminating price trends. During low-price periods, KERAFED's procurement operations and buffer stock schemes should be expanded to avoid distress sales. Strengthening CPCs through targeted training, working capital loans, and marketing support is critical. A state-level e-platform for coconut trade and reforms in trader licensing are also proposed to enhance transparency and competitiveness in the market.