Applied Stochastic Models in Business and Industry最新文献

We propose a real options model with regime shifts to investigate the effect of cutting interest rates on corporate investments when a financial crisis occurs. Cutting interest rates would lower the investment project's hurdle rate. The reduction in hurdle rate is positively related to the magnitude of interest rate cuts and the persistence of the financial crisis. The hurdle rate becomes lower in the financial crisis state because the reduction in interest rate would lower the cost of capital and the opportunity cost of immediate investment. In the numerical analysis of this study, we show that the change in the opportunity cost accounts for most of the change in the hurdle rate. Upon taking into consideration the firm's financing constraints, we find that cutting interest rates accelerates investments for firms with high liquidity. However, for firms with low liquidity, the optimal investment threshold is not affected by the variation in interest rates. Instead, the investments of low-liquidity firms are affected by the change in the friction of credit supply.

Generating realistic vehicle speed trajectories is a crucial component in evaluating vehicle fuel economy and in predictive control of self-driving cars. Traditional generative models rely on Markov chain methods and can produce accurate synthetic trajectories but are subject to the curse of dimensionality. They do not allow to include conditional input variables into the generation process. In this paper, we show how extensions to deep generative models allow accurate and scalable generation. Proposed architectures involve recurrent and feed-forward layers and are trained using adversarial techniques. Our models are shown to perform well on generating vehicle trajectories using a model trained on GPS data from Chicago metropolitan area.

It has been recognized that model risk has an important effect on any risk measurement procedures, particularly when dealing with complex markets and in the presence of a wide range of implemented models. We consider a normalized measure of model risk for the forecast of daily Value-at-Risk, combined with a model selection and an averaging procedure. This allows us to restrict the set of plausible models on a daily basis, making the initial choice of competing models less crucial and then yielding a more reliable assessment of model risk. Using AR-GARCH-type models with different distributions for the innovations, we assess the dynamics of model risk for different financial assets (a stock, an equity index, an exchange rate) and commodities (electricity, crude oil and natural gas) over 15 years.

The comparison of coherent systems in terms of stochastic orders is vital in reliability theory. While there is a considerable amount of literature devoted to comparing systems with homogeneous and independent components, real-world systems often consist of heterogeneous components. Hence, this article aims to investigate systems with heterogeneous and independent components, as well as, those with heterogeneous and dependent components. For this purpose, we consider systems comprise of three components, which are of two different types of components, namely two components of type A and one component of type B. The system's lifetime distribution is represented using the failure signature when the components are independent, which is a function of the component's life distribution. However, when the components are dependent, the system's lifetime distribution is represented using copula and diagonal sections. Additionally, distorted distributions are utilized to enable distribution-free stochastic comparisons. Using these representations, we compare systems with components having proportional reversed hazard rates, in three scenarios: (i) when components are heterogeneous and independent; (ii) when components are heterogeneous and dependent; and finally, (iii) comparing systems with homogeneous and independent components with those that have heterogeneous components. To illustrate the applicability of these results, we provide some examples and applications.

Commodity price volatility is a major source of instability in those countries that are primarily commodity-dependent and has a negative impact, especially on economic growth. With this premise, commodities represent an effective financial exchange tool that nowadays finds relevance in being involved in the processes inherent to environmental sustainability. This work focus on raw materials and their demand, connected with the need for a transition towards the Circular Economy, as part of a strategy to address commodity supply disruptions. It presupposes changes in the mentality and behavior of companies in the various economic sectors. A crucial issue debated in the literature concerns whether or not the size of the companies favors their attitude towards Circular Economy. We propose a nonparametric method to test the effect of firm size on their propensity to undertake Circular Economy activities. Considering $��k�$ of such activities, this propensity is a multidimensional concept and it can be represented by a $��k�$-dimensional vector of proportions. Each element of the vector represents the share of companies of the population under study that implement a specific Circular Economy activity. The main difficulty of such a multivariate testing problem, together with the multidimensional nature of the dichotomous response, is the one-sided type alternative, which is a stochastic dominance for multidimensional binary variables. A Monte Carlo simulation study proves the good power behavior of the proposed solution, based on a nonparametric approach. Case studies related to Italian small and medium enterprises in some strategic sectors are also addressed.

We propose a reinforcement learning (RL) algorithm for generating a trading strategy in a realistic setting, that includes transaction costs and factors driving the asset dynamics. We benchmark our algorithm against the analytical optimal solution, available when factors are linear and transaction costs are quadratic, showing that RL is able to mimic the optimal strategy. Then we consider a more realistic setting, including non-linear dynamics, that better describes the WTI spot prices time series. For these more general dynamics, an optimal strategy is not known and RL becomes a viable alternative. We show that on synthetic data generated from WTI spot prices, the RL agent outperforms a trader that linearizes the model to apply the theoretical optimal strategy.

Albeit not equivalent, in many applications the gamma and the inverse Gaussian processes are treated as if they were. This circumstance makes the misspecification problem of these models interesting and important, especially when data are affected by measurement errors, since noisy/perturbed data do not allow to verify whether the selected model is actually able to adequately fit the real (hidden) degradation process. Motivated by the above considerations, in this paper we conduct a large Monte Carlo study to evaluate whether and how the presence of measurement errors affects this misspecification issue. The study is performed considering as reference models a perturbed gamma process recently proposed in the literature and a new perturbed inverse Gaussian process that share the same non-Gaussian distributed error term. As an alternative option, we also analyze the more classical case where the error term is Gaussian distributed. We consider both the situation where the true model is the perturbed gamma and the one where it is the perturbed inverse Gaussian. Model parameters are estimated from perturbed data using the maximum likelihood method. Estimates are retrieved by using a new sequential Monte Carlo EM algorithm, which use allows to hugely mitigate the severe numerical issues posed by the direct maximization of the likelihood. The risk of incurring in a misspecification is evaluated as percentage of times the Akaike information criterion leads to select the wrong model. The severity of a misspecification is evaluated in terms of its impact on maximum likelihood estimate of the mean remaining useful life.

One of the most challenging aspects for managers when building a forecasting system is choosing how to aggregate the data at different levels. This is frequently done without the manager knowing how these choices can compromise the system's accuracy. This article illustrates these compromises by comparing different structures and aggregation criteria. Our article proposes and empirically tests a framework on how to build a coherent and more accurate forecasting system. The framework's first phase compares different time series forecasting methods, including statistical, “standard” machine learning, and deep learning. Results show that one of the statistical methods (autoregressive integrated moving average, or, for short, ARIMA) outperforms machine and deep learning methods. The second phase compares different combinations of aggregation criteria, structures of the forecasting system, and coherent forecast methods (i.e., adjustments to the forecasts at different levels of aggregation). The results show that using different criteria and structures indeed impacts predictions' accuracy. When it is necessary to disaggregate the forecast, our results show that it is best to add more information in a grouped structure, adjusted by a bottom-up method. This combination provides the best performance, that is, the lowest mean absolute-scaled error (MASE) in most nodes, compared to the other structures and coherent forecast methods used. The results also suggest that aggregating the time series further by geographical regions is essential to improve accuracy when forecasting products' and channels' sales.