Editor’s Note
Dr. Rajib Bhattacharya, teaches Finance at Globsyn Business School – one of Kolkata’s best colleges for MBA in international business. Dr. Bhattacharya has over 14 years of experience in middle and senior level positions in various industries including the banking sector. In his academic career extending over 11 years, he has been associated with reputed business schools as faculty member, paper setter and examiner. He had also been a guest faculty at Universitas Pelita Harapan, Jakarta, Indonesia. He has provided corporate training to middle and senior level executives of many MNCs. He has many research papers (including a best-paper award winning paper) to his credit, published in peer-reviewed indexed journals and presented in national & international conferences including some at IITs and IIMs. He is a member of The Institute of Cost Accountants of India, All India Management Association, Operational Research Society of India and Analytics Society of India.
Extreme Value Theory (EVT) is a statistical branch that focuses on the analysis and modeling of rare or extreme events that occur infrequently but may have significant consequences. It provides tools and techniques to understand and quantify the tail behavior of a distribution, which is beyond the range of traditional statistical methods. EVT finds its applications in various fields, including finance, insurance, engineering, environmental science, and meteorology.
The central concept of EVT is the ‘Extreme Value’ or ‘Tail’ of a distribution. EVT assumes that the extreme events are generated by a different process compared to the bulk of the data. EVT distinguishes between two types of extreme events: the block-maxima approach, which considers the maximum value of a block of data over a fixed period, and the peak-over-threshold approach, which considers values exceeding a high threshold.
One of the fundamental results of EVT is the ‘Extreme Value Theorem’ or the ‘Fisher-Tippett-Gnedenko Theorem’, which states that the extreme values of a distribution converge to one of three limiting distributions: the Gumbel distribution for the block-maxima approach, the Generalized Extreme Value (GEV) distribution for the peak-over-threshold approach, and the Generalized Pareto Distribution (GPD) for excesses over a high threshold.
EVT provides several useful tools for estimating the parameters of these extreme value distributions, such as the maximum likelihood estimation, the method of moments, and the probability weighted moments. These methods allow practitioners to estimate the tail parameters of a distribution, which are crucial for understanding the tail behavior and quantifying the risk of extreme events.
Application Areas of EVT
Extreme Value Theory (EVT) has found numerous applications in various fields due to its ability to model and analyze rare or extreme events. The major application of EVT is in any domain areas of Risk Management, where it is used to estimate the tail risk or the likelihood of extreme events, such as financial market crashes or natural disasters, that may have severe consequences. EVT also finds applications in reliability analysis, where it is used to estimate the probability of failure of critical systems or components.
In addition, EVT is used in environmental science to model extreme weather events, such as hurricanes or floods, for planning and mitigation purposes. Some application areas where EVT is used frequently, are appended below:
Finance and Insurance: EVT is widely used in the field of finance and insurance for proactive risk management and estimation of tail risk. EVT can help financial institutions estimate the likelihood of extreme events such as market crashes, stock market declines, or rare events in insurance claims, which are crucial for pricing financial products, managing investment portfolios, and setting insurance premiums. EVT has been applied in modeling extreme events in stock returns, currency exchange rates, credit risk, and insurance losses, among others.
Environmental Science: EVT is used in environmental science to model and predict extreme weather events, such as hurricanes, floods, and droughts, which are critical for disaster risk reduction and management. EVT helps in estimating the intensity, frequency, and duration of extreme weather events, providing valuable information for infrastructure planning, resource management, and policy-making in the fields of hydrology, climatology, and environmental engineering.
Engineering: EVT is employed in engineering for analyzing and designing structures or systems that are exposed to extreme loads, such as wind turbines, bridges, dams, and offshore platforms. EVT can help estimate the probability of failure of critical components or systems under extreme conditions, which is crucial for safety and reliability assessments. EVT has also been used in the field of telecommunications to model extreme traffic loads and estimate capacity requirements for communication networks.
Energy and Utilities: EVT is used in the energy and utilities sector for estimating extreme events related to energy production, transmission, and distribution. EVT can help estimate the likelihood of extreme events such as electricity demand peaks, power grid failures, or extreme weather conditions affecting energy infrastructure. This information is essential for planning and managing energy resources, optimizing energy production, and ensuring the reliability of energy supply.
Environmental Risk Assessment: EVT is employed in environmental risk assessment to model and estimate the probability of extreme events related to environmental hazards such as earthquakes, landslides, forest fires, and pollution incidents. EVT can help estimate the likelihood and severity of extreme events, assess their potential impacts on the environment and human health, and inform decision-making in risk mitigation, emergency preparedness, and environmental policy.
Transportation: EVT is used in transportation for analyzing and managing risks associated with extreme events such as traffic accidents, road or rail disruptions, and extreme weather conditions that can impact transportation systems. EVT can help estimate the likelihood of rare events in transportation networks, optimize traffic management strategies, and inform infrastructure planning for robust and resilient transportation systems.
Sports Analytics: EVT is used in sports analytics to model and predict extreme events in sports performance, such as record-breaking performances, outlier performances by athletes, and rare events in sports outcomes. EVT can help in estimating the probabilities of extreme events in sports, analyzing performance trends, and informing sports strategy and training programs.
Health and Medicine: EVT is employed in health and medicine for analyzing and predicting rare events in health outcomes, such as disease outbreaks, medical emergencies, or adverse drug reactions. EVT can help estimate the likelihood of rare events in health data, assess the impacts of extreme events on public health, and inform health policy and intervention strategies.
EVT provides valuable tools and techniques for estimating the probabilities of extreme events and taking informed decision-making in proactive risk management, planning, and policy formulation. EVT plays a critical role in helping practitioners better understand the behavior of extreme events and make informed decisions to mitigate risks and improve resilience in situations when these extreme events happen.
Globsyn Business School prepares its students to face, right from day one, the complex business environment which, from being a VUCA (Volatile, Uncertain, Complex, Ambiguous) one, to become VUCCA (Volatile, Uncertain, Complex, Chaotic, Ambiguous), and is currently a BANI (Brittle, Anxious, Nonlinear, Incomprehensible).
Accordingly, the students at GBS are exposed to the latest Risk Management tools like EVT so that they can be effective and efficient managers in their careers in the corporate world.
Dr. Rajib Bhattacharya
Faculty – Finance
Globsyn Business School