Risk-neutral Approaches Research Articles

An assessment of model risk in pricing wind derivatives

AbstractWind derivatives are financial instruments designed to mitigate losses caused by adverse wind conditions. With the rapid growth of wind power capacity due to efforts to reduce carbon emissions, the demand for wind derivatives to manage uncertainty in wind power production is expected to increase. However, existing wind derivative literature often assumes normally distributed wind speed, despite the presence of skewness and leptokurtosis in historical wind speed data. This paper investigates how the misspecification of wind speed models affects wind derivative prices and proposes the use of the generalized hyperbolic distribution to account for non-normality. The study develops risk-neutral approaches for pricing wind derivatives using the conditional Esscher transform, which can accommodate stochastic processes with any distribution, provided the moment-generating function exists. The analysis demonstrates that model risk varies depending on the choice of the underlying index and the derivative’s payoff structure. Therefore, caution should be exercised when choosing wind speed models. Essentially, model risk cannot be ignored in pricing wind speed derivatives.

A Risk-Sensitive Approach to Policy Optimization

Standard deep reinforcement learning (DRL) aims to maximize expected reward, considering collected experiences equally in formulating a policy. This differs from human decision-making, where gains and losses are valued differently and outlying outcomes are given increased consideration. It also fails to capitalize on opportunities to improve safety and/or performance through the incorporation of distributional context. Several approaches to distributional DRL have been investigated, with one popular strategy being to evaluate the projected distribution of returns for possible actions. We propose a more direct approach whereby risk-sensitive objectives, specified in terms of the cumulative distribution function (CDF) of the distribution of full-episode rewards, are optimized. This approach allows for outcomes to be weighed based on relative quality, can be used for both continuous and discrete action spaces, and may naturally be applied in both constrained and unconstrained settings. We show how to compute an asymptotically consistent estimate of the policy gradient for a broad class of risk-sensitive objectives via sampling, subsequently incorporating variance reduction and regularization measures to facilitate effective on-policy learning. We then demonstrate that the use of moderately "pessimistic" risk profiles, which emphasize scenarios where the agent performs poorly, leads to enhanced exploration and a continual focus on addressing deficiencies. We test the approach using different risk profiles in six OpenAI Safety Gym environments, comparing to state of the art on-policy methods. Without cost constraints, we find that pessimistic risk profiles can be used to reduce cost while improving total reward accumulation. With cost constraints, they are seen to provide higher positive rewards than risk-neutral approaches at the prescribed allowable cost.

Single-stage gradient-based stellarator coil design: stochastic optimization

We extend the single-stage stellarator coil design approach for quasi-symmetry on axis from (Giuliani et al 2020) to additionally take into account coil manufacturing errors. By modeling coil errors independently from the coil discretization, we have the flexibility to consider realistic forms of coil errors. The corresponding stochastic optimization problems are formulated using a risk-neutral approach and risk-averse approaches. We present an efficient, gradient-based descent algorithm which relies on analytical derivatives to solve these problems. In a comprehensive numerical study, we compare the coil designs resulting from deterministic and risk-neutral stochastic optimization and find that the risk-neutral formulation results in more robust configurations and reduces the number of local minima of the optimization problem. We also compare deterministic and risk-neutral approaches in terms of quasi-symmetry on and away from the magnetic axis, and in terms of the confinement of particles released close to the axis. Finally, we show that for the optimization problems we consider, a risk-averse objective using the conditional value-at-risk leads to results which are similar to the risk-neutral objective.

Longevity Modelling and Pricing under a Dependent Multi-Cohort Framework

We propose a multi-cohort model that is able to capture the mortality correlation between different cohorts. The model is based on the Hull and White process to which we incorporate inter-generational risk factors, by modifying its stochastic part. We provide a pricing framework for a new survival forward contract under the Cost of Capital, risk-neutral and Sharpe approaches, allowing to cover the global multi-cohort longevity risk. We give numerical illustrations for Belgian cohorts, and we compute the price of the longevity derivative under the proposed methods, for different correlation levels.

Hotel room allocation with multiple sales channels: a perspective of minimax regret

ABSTRACT We consider a room allocation problem for a hotel providing room reservation service through multiple channels. Traditional room allocation models usually assume hotels can forecast the characteristics of demand function and/or arrival process to some extent. But in practice, such a restrictive assumption is possibly unrealistic, and thus the corresponding risk-neutral approaches are debatable. The main contribution of this paper is that we formulate the hotel’s room allocation problem without any assumption on demand from the perspective of minimax regret and derive an adaptive nested limits strategy which allows higher profitable channels having access to all rooms reserved for lower profitable channels. Our results show that customers arrive in a low-before-high manner in the worst-case scenarios. On this basis, we further prove the lower bound of maximum regret for any strategy and develop an optimal strategy matching the lower bound. Comparison with other common strategies is also discussed.

Time-consistent, risk-averse dynamic pricing

Many industries use dynamic pricing on an operational level to maximize revenue from selling a fixed capacity over a finite horizon. Classical risk-neutral approaches do not accommodate the risk aversion often encountered in practice. When risk aversion is considered, time-consistency becomes an important issue. In this paper, we use a dynamic coherent risk-measure to ensure that decisions are actually implemented and only depend on states that may realize in the future. In particular, we use the risk measure Conditional Value-at-Risk (CVaR), which recently became popular in areas like finance, energy or supply chain management.A result is that the risk-averse dynamic pricing problem can be transformed to a classical, risk-neutral problem. To do so, a surprisingly simple modification of the selling probabilities suffices. Thus, all structural properties carry over. Moreover, we show that the risk-averse and the risk-neutral solution of the original problem are proportional under certain conditions, that is, their optimal decision variable and objective values are proportional, respectively. In a small numerical study, we evaluate the risk vs. revenue trade-off and compare the new approach with existing approaches from literature.This has straightforward implications for practice. On the one hand, it shows that existing dynamic pricing algorithms and systems can be kept in place and easily incorporate risk aversion. On the other hand, our results help to understand many risk-averse decision makers who often use “conservative” estimates of selling probabilities or discount optimal prices.

A risk-averse stochastic program for integrated system design and preventive maintenance planning

The failure of high-consequence systems, such as highspeed railways, can result in a series of severe damages. Due to the volatility of real circumstances, stochastic optimization methods are needed to aid decisions on reliability design and maintenance for the high-consequence systems. Traditionally, risk-neutral approaches are used by considering the expectation of random variables as a preference criterion. The risk-neutral approaches may achieve the solutions that are good in the long run but do not control poor results under certain realizations of random variables. From a perspective of risk analysis, such solutions are not acceptable for high-consequence systems. To address this issue, this paper uses the conditional value at risk (CVaR) to more properly account for some of the worst realizations of random future usage scenarios, and then proposes a risk-averse two-stage stochastic programming model to simultaneously determine the numbers of components in each subsystem and preventive maintenance time intervals for the high-consequence systems that are exposed to uncertain future usage stresses. The proposed stochastic programming model can be converted to a nonconvex mixed-integer nonlinear programming (MINLP) model. To solve the model, we derive the analytical properties of the recourse function and the closed form of CVaR and then design a decomposition algorithm. Numerical examples demonstrate the proposed risk-averse stochastic approach and the effectiveness of incorporating the CVaR in modeling. The results show the research problem significantly benefits from the proposed approach. Furthermore, the robustness of the optimal system design and maintenance plan under different profiles of future usage scenarios is addressed.

Developing a Risk-Based Approach for American Basket Option Pricing

Options are one of the important financial contracts for reducing the risk of investors. Many active practitioners in the financial markets really believe that mispricing or incorrect valuation of these securities would be the main reason of collapse of some financial institutions. The complexity of option pricing/valuation, especially in the case of American basket options, as high dimensional options, has motivated many researchers to develop numerical and simulation-based models. In this paper, a new simulation-based approach for pricing/valuation of American basket option with risk consideration is proposed. Having the prices obtained through Longstaff–Schwartz methodology, which is based on Approximate Dynamic Programming as a risk-neutral approach, we propose a new approach for pricing the American basket option according to the worst-case (pessimistic/risk-averse) and the best-case (optimistic/risk-taking) scenarios. Furthermore, for scenarios generation, we use a Monte Carlo simulation technique using a t-student copula-GARCH method and Extreme Value Theory to handle the nonlinearity of dependencies between variables. To verify the computational efficiency and the accuracy of the proposed methodology, we compare the results of prices obtained through the proposed models with those achieved through the Monte Carlo simulation and the method developed by Ju for European basket options. Moreover, the developed models are tested using out-sample scenarios to explore what would happen if investors bought the option with the obtained prices through three different strategies: risk-averse, risk-neutral, and risk-taking approaches.

Optimizing conditional value-at-risk in dynamic pricing

Many industries use dynamic pricing on an operational level to maximize revenue from selling a fixed capacity over a finite horizon. Classical risk-neutral approaches do not accommodate the risk aversion often encountered in practice. We add to the scarce literature on risk aversion by considering the risk measure conditional value-at-risk (CVaR), which recently became popular in areas like finance, energy, or supply chain management. A key aspect of this paper is selling a single unit of capacity, which is highly relevant in, for example, the real estate market. We analytically derive the optimal policy and obtain structural results. The most important managerial implication is that the risk-averse optimal price is constant over large parts of the selling horizon, whereas the price continuously declines in the standard setting of risk-neutral dynamic pricing. This offers a completely new explanation for the price-setting behavior often observed in practice. For arbitrary capacity, we develop two algorithms to efficiently compute the value function and evaluate them in a numerical study. Our results show that applying a risk-averse policy, even a static one, often yields a higher CVaR than applying a dynamic, but risk-neutral, policy.

Practical decision rules for risk-averse revenue management using simulation-based optimization

In practice, human-decision makers often feel uncomfortable with the risk-neutral revenue management systems’ output. Reasons include a low number of repetitions of similar events, a critical impact of the achieved revenue for economic survival, or simply business constraints imposed by management. However, solving capacity control problems is a challenging task for many risk measures and the approaches are often not compatible with existing software systems. In this paper, we propose a flexible framework for risk-averse capacity control under customer choice behavior. Existing risk-neutral decision rules are augmented by the integration of adjustable parameters. Our key idea is the application of simulation-based optimization (SBO) to calibrate these parameters. This allows to easily tailor the resulting capacity control mechanism to almost every risk measure and customer choice behavior. In an extensive simulation study, we analyze the impact of our approach on expected utility, conditional value-at-risk (CVaR), and expected value. The results show a superior performance in comparison to risk-neutral approaches from the literature.

Incorporating risk measures in closed-loop supply chain network design

This paper considers a location-allocation problem in a closed-loop supply chain (CLSC) with two extensions: first, demand and prices of new and return products are regarded as non-deterministic parameters and second, the objective function is developed from expected profit to three types of mean-risk ones. Indeed, design and planning an integrated CLSC in real-world volatile markets is an important and necessary issue. Further, risk-neutral approaches, which are considered expected values, are not efficient for such uncertain conditions. Hence, this paper, copes with the design and planning problem of a CLSC in a two-stage stochastic structure. Besides, risk criteria are considered through using three types of popular and well-behaved risk measures: mean absolute deviation, value at risk and conditional value at risk (CVaR). Consequently, three types of mean-risk models are developed as objective functions and decision-making procedures are undertaken based on the expected values and risk adversity criteria. Finally, performances of the developed mean-risk models are evaluated in various aspects. Results reveal that the inefficiencies of risk-neutral approaches can be overcome. In addition, in terms of quality of solutions, the acceptability of CVaR is proved when it is compared to other risk measures.

Variational inequalities for solving possibilistic risk-averse electricity market equilibrium

It is widely known and accepted that Nash equilibrium suitably models agents' behavior in electricity markets, since it is coherent with the common sense of their simultaneous profits maximisation. In the literature, these approaches are usually addressed using deterministic representations, despite the fact that electricity markets are highly conditioned by the uncertainty in demand or in agents' bidding strategies. Only some equilibrium-modelling approaches under uncertainty can be found in the literature, most of them using probability distributions. However, probability approaches may lead to very complex formulations and generally require restrictive assumptions (such as normality or independence) that can hardly be verified in real complex problems. A conjectured-price-response equilibrium model that uses LR-possibility distributions to represent the uncertainty of the residual demand curves faced by the participant agents is proposed. Modelling the risk-aversion attitudes of the agents, the resulting possibilistic equilibrium is transformed into a simplified deterministic one, which is solved with a new globally convergent algorithm for variational inequalities problems. Some interesting results for a real-size electricity system show the robustness of this new approach when compared with other risk-neutral approaches.