Tales from tails: On the empirical distributions of forecasting errors and their implication to risk

When evaluating the performances of time series extrapolation methods, both researchers and practitioners typically focus on the average or median performance according to some specific error metric, such as the absolute error or the absolute percentage error. However, from a risk-assessment point of view, it is far more important to evaluate the distributions of such errors, and especially their tails. For instance, a lack of normality and symmetry in error distributions can have significant implications for decision making, such as in stock control. Moreover, frequently these distributions can only be constructed empirically, as they may be the result of a computationally-intensive non-parametric approach, such as an artificial neural network. This study proposes an approach for evaluating the empirical distributions of forecasting methods and uses it to assess eleven popular time series extrapolation approaches across two different datasets (M3 and ForeDeCk). The results highlight some very interesting tales from the tails.     

A strategic planning model for the railway system accident rescue problem

The present study examines the location of emergency rescue problems for urban ambulance and railway emergency systems. The proposed model considers probabilistic rescue demand, independent busy fractions of ambulances, and the corresponding risk levels in railway segments. We formulate the proposed model using fuzzy multi-objective programming and solve it using a generic algorithm and a non-dominated sorting genetic algorithm-II. Computation results are analyzed by applying the model to a real-world Taiwan railway system. Analytical results demonstrate that a proper adjustment of the rescue resource location improves rescue effectiveness for railway rescue and urban medical service demand.     

Are impatient farmers more risk-averse? Evidence from a lab-in-the-field experiment in rural Uganda

Based on data from a field-experiment in rural Uganda, we show that impatient farmers are more risk-averse than patient farmers. We relied on a simplified version of the Convex Time Budget (CTB) method to elicit farmers’ time preferences and on an independent method for eliciting their risk-preferences. We report two important findings. First, we show that our simplified CTB method applied to farmers from Uganda replicates the key findings of Andreoni and Sprenger’s lab experiments that involved student subjects. Second, we establish the existence of a negative correlation between risk tolerance and impatience, based on two independent measures.     

Risk shifting in the US banking system: An empirical analysis

This paper contributes to the empirical literature on risk shifting. It proposes a method to find out whether risk shifting is present in the banking industry and, if so, what type. The type of risk shifting depends on the group of debt holders to whom risk is shifted. We apply this method to the US banking sector in 1998–2011. To study the relationship between risk shifting and the 2008 crisis, the sample is also split into pre-crisis, crisis, and post-crisis periods. Our results suggest that the same type of risk shifting is present in the entire sample and in the pre-crisis and crisis subsamples. We find no evidence of risk shifting after the crisis. Furthermore, holding capital buffers seems to disincentivize risk shifting. This finding appears to provide support for the conservative buffer included in Basel III.     

Perceived risk and insurance decision taking for small losses

Abstract

Little systematic research has investigated differences in expressed attitude as a function of the manner in which probability information is communicated to a decision maker. The purpose of this paper is to investigate differences in expressed attitude when insurance coverage is introduced in a known-risk situation (the probability of loss is known), an uncertain situation (there is no prior information on the probability of loss) or an ambiguous (the information provided is vague). The experiments reported in this paper have been developed and tested in the classroom with undergraduate students. Unlike the vast majority of previous work dealing with lotteries and laboratory gambles, this study examine the behavior of people when facing a purchase decision on a well-known consumer good, i.e. a bottle of wine. The comparative results provide some evidence on the risk-taking behavior of consumers for small losses. Within an insurance context, people prefer the more familiar option of a known-risk situation and contrary to expectations, the results do not provide support to ambiguity aversion but to ambiguity seeking.     

Diversified behavioral portfolio as an alternative to Modern Portfolio Theory

The traditional mean–variance approach has been complemented by alternative theories that use risk measures different from standard deviation of returns or involve additional distributional features of returns like skewness and kurtosis. We propose a portfolio choice model that combines different distributional characteristics of the returns in the decision-making making process, considering preferences of investors which are modeled as non-statistical uncertainties of investors using fuzzy theory. We use 20 stocks of the S&P500 from January 2013 to December 2017. We assess the obtained portfolios’ performance, and the diversified behavioral portfolios outperform than the mean–variance portfolio. This methodological proposal can be seen as a strong managerial tool to make investment portfolio decisions.     

Multinational enterprises’ risk mitigation strategies in emerging markets: A political coalition perspective

This paper contributes to the IB literature by investigating MNEs’ risk mitigation strategies in emerging markets. Drawing on institutional perspectives and March’s theoretical concept of forming a ‘political coalition’, we propose that risk mitigation cannot be limited to passive compliance and/or demonstrating good corporate behavior, but should extend to collective efforts by building a political coalition and working with key stakeholders to manage potential risk and obtain favorable outcomes in complex institutional environments. In considering MNEs in contemporary China and India, we offer different types of risk mitigation strategies under various institutional contexts and a framework for future research.     

Real estate prices and systemic banking crises

The collapse of real estate prices has historically jeopardized banking stability and triggered systemic banking crises. This paper studies risk contagion in a banking system in real estate price shock by adopting complex network theory. Modelling the real estate-related asset as a common exposure of banks to the real estate market, we propose a model that incorporates two main risk contagion channels, i.e., the financial network and asset fire sales, and reveal how the real estate price shock is transmitted and propagated across banks. We demonstrate that banking stability is highly sensitive to the real estate price shock. Moreover, due to the particularly low liquidity of the real estate market, the asset fire-sales of real estate assets overwhelms the financial network, playing the dominant role in risk contagion. Our model can be adopted by regulators to conduct stress testing and to forge effective risk management strategies.     

Empirical evidence of extreme dependence and contagion risk between main cryptocurrencies

This study examines the dependence and contagion risk between Bitcoin (BTC), Litecoin (LTC) and Ripple (XRP) using non-parametric mixture copulas (developed by Zimmer, 2012) and recently proposed methods of full-range tail dependence copulas (advanced by Hua, 2017, Su and Hua, 2017), for the period from 04-08-2013 to 17-06-2018. The Chi-plots and Kendall plots results show heavy tail dependence between each pairs of the cryptocurrencies. Evidence from the mixture copula indicates that for the BTC-LTC pair the upper-tail dependence is both stronger and more prevalent, while for the other pairs of cryptocurrencies the lower-tail dependence is very strong and more prevalent. However, the results of the full-range tail dependence copulas reveal a strong and prevalent upper and lower-tail dependence of each pairs of cryptocurrencies. These results provide evidence of significant risk contagion among price returns of major cryptocurrencies, both in bull and bear markets.     

Supply chain risk analysis: a shipbuilding industry case

A lack of empirical studies on Supply Chain Risk Management (SCRM) is observed, especially in the industries of large projects. Therefore, this paper conducts a risk analysis on the Brazilian Shipbuilding Industry by generating an initial risk profile. A panel was conducted with executives from major shipyards, and the primary risks were identified, assessed, and evaluated. Findings corroborate the academic literature, reinforcing that companies recognise the importance of SCRM, but the concept is incipient and far from being successfully implemented in real life. Despite the fact that the shipyards identified important risks, the focus was on internal factors. Because the industry was exposed to risks from the external environment, no risk mitigation actions existed, resulting in the bankruptcy of many shipbuilders. This research is an initial step toward investigating SCRM techniques, offering academics a novel empirical approach that can serve as a systematic risk management tool for supply chain planning.