Optimal fisheries investment

This thesis explores problems of optimal investment arising in fisheries and other renewable resource industries. In such industries, two simultaneous investment problems must be addressed: investment in the resource stock (the biomass) and investment in the capital stock (harvesting capacity). Each of these investment problems faces a key complication; investment in the resource is constrained by the natural population dynamics, while in many cases investment in the physical capital stock suffers from irreversibility, since capital used in natural resource industries is often non-malleable. In addition, all investment decisions must be made within an uncertain environment; full information is never available. Building upon the work of Clark, Clarke and Munro (Econometrica, 1979), we develop a two-state two-control model which incorporates investment delays and stochastic resource fluctuations within a seasonal (discrete-time) framework. A dynamic programming approach is used to analyse the model heuristically and to obtain numerical results, beginning with a study of the ideal deterministic case and proceeding to a full stochastic analysis. The key assumption of irreversible investment is maintained throughout the thesis. We have examined the qualitative and quantitative effects on optimal investment strategies of several economic and ecological factors: (i) delays in investment, (ii) population dynamics parameters, (iii) selling price, (iv) capital cost and operating cost, (v) depreciation rate, (vi) discount rate, and (vii) the level of uncertainty in the resource stock. We have found that the key cost parameter for the investment problem is the ratio of unit capital costs to unit operating costs. Depreciation can play a rather counter-intuitive role; in some circumstances optimal investment levels can increase with the depreciation rate, contrary to the usual treatment of depreciation as an additional cost of capital. The introduction of uncertainty in the form of stochastic resource fluctuations can substantially change the optimal investment policy, but this tends to have little effect on the value of the fishery. We analyse the factors which determine the role of randomness in optimal fisheries investment, and discuss in some detail the implications for management. Solution of the stochastic optimization problem studied here requires the use of rather complicated numerical methods, which are described in detail in the thesis. These methods are quite general, and should prove useful in analysing other related stochastic models.Science, Faculty ofMathematics, Department ofGraduat