Production risk and technical efficiency of Thai Koi (Anabas testudineus Bloch) farming in some selected areas of Bangladesh

Thai koi production technology is relatively a new component in the aquaculture industry in Bangladesh. However, despite its impressive growth during the initial years, the industry is now experiencing volatility in terms of year to year fluctuation in production. The sharp fluctuations in production make koi farming a risky venture. Thus koi farming, as a spontaneous choice has become questionable. The overall objective of this study is to examine the production risk and technical efficiency (TE)of Thai koi farming in selected areas of Bangladesh. Specifically it aims at:(i)identifying the nature and extent of risk associated with the use of inputs; (ii) estimating the farm specific as well as mean TE and identifying factors affecting technical inefficiency (TIE), and (iv) assessing how feed type effects on productivity. The study utilized a total of 149cross-sectional data collected randomly from the three sub-districts of Mymensingh district in Bangladesh. The study employed Kumbhakar’s (2002) approach using stochastic frontier model (SFM) to estimate the parameters of mean production, variance and inefficiency functions in a single framework. Influence of different feed types on productivity growth was investigated using Cobb-Douglas(C-D) production function and Propensity Score Matching (PSM) methods. The findings revealed that among the productive variables, feed was the most important input influencing Thai koi farming. Production of koi was characterized by decreasing returns to scale. The estimated risk functions showed that the zeolite and pesticide are risk-reducing, while labor, feed, fingerlings and salt appeared to be riskincreasing inputs. Results of the inefficiency model indicated that labor and pond area significantly reduced the TIE. Feed, fingerlings and lime appeared to influence TE. Education of farmers’ increased TE, while experience, training and water change affected TE negatively. The mean TE scores were 0.73 and 0.96 for the flexible risk estimate and conventional SFM estimate, respectively. The estimated TE using the flexible risk framework indicates that 27% of the potential output could be recovered by removing inefficiency. The analyses suggest that TE values are exaggerated when production variance was not included in the model. Thus, the study advocates that production-TE modelling should be done using flexible risk framework without being illusive with high TE values estimated through conventional risk-free framework. Floating feed showed better performance in terms of the mean daily weight gain of fish, feed conversion ratios and mean weight of table fish with less culture days as opposed to sinking and mixed feeds. The net returns and benefit cost ratio were also the highest with the floating feed. The C-D regression and PSM results confirm that farmers using floating feed produced higher quantity of fish per unit compared to those using the other feed types. However, farms using the floating feed need more investment to improve their returns. The overall findings suggest that there was a lack of proper understanding of the technology practices. The farmers could become more efficient by adjusting input allocation with the existing technology and resources. Further development of this agri-business requires for the reduction in major variable costs and ensures its sustainability. In particular, the development of low-cost feeds would help to secure farmer’s profit margins at a competitive level. Finally, the outlook for the Thai koi culture could be very positive in meeting the future demand of fish. Hence, rectifying the problems and constraints faced at the farms level would enhance the sector’s contribution to food and nutrition security in the area