This paper develops a modeling framework for biomass futures contracts by connecting biological fish growth dynamics with financial risk management tools. We model the salmon farming process in the presence of environmental and biological risks using a multi-factor stochastic model and capture optimal harvesting behavior within a real options approach. The corresponding dynamic problem is solved via the Longstaff-Schwartz algorithm. By incorporating biomass-based derivatives, our framework enables farmers to hedge against production risk arising from fluctuating water temperatures, mortality and ecological volatility. Simulation results show that while biomass futures may reduce expected profits under ideal growing conditions, they offer significant value under thermal stress and extreme scenarios by stabilizing income and reducing downside risk. From a methodological perspective, we employ machine learning tools, e.g., bagged decision tree ensemble, to estimate the optimal harvesting boundary, which is a two-dimensional hypersurface within the three-dimensional state space. The proposed approach demonstrates how linking biological systems with financial instruments can enhance decision-making, improve income stability, and support sustainability in aquaculture. Although focused on farmed Atlantic salmon, the framework is adaptable to other species and environmental settings.

📄 View full paper on SSRN

We consider large-scale commodity project investments within a real options framework where, in addition to risk, the investor faces true uncertainty, henceforth referred to as ambiguity. Within this framework, we use a new Monte Carlo based method to determine the optimal exercise time and the value of the real option representing the worth of an investment project. Model ambiguity arises due to uncertain parameter values in the model, and its severity is captured via confidence intervals of parameter estimates. This is the situation in most real-world projects; however, in academic work, it seems largely ignored. The classical Monte Carlo approach of Longstaff and Schwartz is unable to deal with such a setup. Our proposed approach, therefore, provides a significant extension of the computational toolbox for real options. Our approach is based on the insight that the value of a real option has an interpretation as the solution of a reflected backward stochastic differential equation (RBSDE), and that the part of the dynamic optimization process which depends on model ambiguity translates to a point-wise optimization of the driver function of the related RBSDE. This insight is then combined with a computational methodology for solving RBSDEs known as the Stratified Regression One-step Forward Dynamic Programming algorithm. To demonstrate the applicability of our framework, we consider a standard model of aquaculture investment and quantify the ambiguity premium due to parameter uncertainty. Our numerical results demonstrate that the ambiguity premium is substantial and that an increase in uncertainty leads to an earlier optimal harvesting time.

📄 View full paper on SSRN

This paper studies the timing of firm investment under uncertainty about taxation policy. Under such uncertainty, firms face a trade-off between investing early, thus avoiding the new tax but relinquishing some depreciation benefits on capital expenditures, and investing later, at which point the tax may already be in place. This tradeoff influences not only the firm’s optimal investment timing but also its capacity decisions and overall expected welfare, highlighting how tax policy uncertainty, tax magnitude and depreciation rules directly affects both capacity choices and welfare outcomes. These insights are particularly relevant in the context of sustainability transitions, where governments are increasingly seeking to internalize negative externalities, such as emissions and resource depletion, by introducing new taxes. Firms operating in natural resource industries, for instance, often must weigh the potential costs of upcoming taxes against the benefits of earlier investments.

The financial sector is increasingly acknowledging sustainability factors as material risks for corporations. This is particularly relevant for the economic activities that are most reliant on ecosystem services, such as food production, which are at the same time the most vulnerable to and the major drivers of the ecological crisis.  Salmon farming, one of the fastest-growing food production systems worldwide, generates substantial negative impacts on marine ecosystems that also affect the financial performance of producers. Yet, while previous research has examined the factors affecting the profitability of the industry, most studies have focused on financial rather than environmental drivers. In this paper, we fill this gap by analysing the relationship between biological impacts and profitability in Norwegian salmon aquaculture. Using data from the 36 largest producers, we apply a dynamic panel approach based on the system generalised method of moments to estimate how the return on assets is affected by biological variables, such as sea lice, escapes, diseases outbreaks, bottom conditions, and water temperature.   Our results show that, with the exception of escapes, these factors significantly reduce profitability, indicating that most biological impacts are financially material. These findings suggest that reducing biological impacts is not only environmentally beneficial but also economically advantageous, while the lack of a measurable effect of escapes highlights the need for improved monitoring and reporting.