CRAFT Center Current Research

Magnifying glass focusing on financial data

SHIFT: A Test Bed for Risk Assessment & Regulatory Compliance

SHIFT is not only a market replay tool but also allows trading algorithms to interact with each other. Trading agents implemented in the system drive the resulting price process. The system may be used for training and assessment of risk.

Learn More About This Project

Explainable AI with Attention Networks for Credit/Loan Decisions

Reasonably explain the decisions suggested by the AI algorithms; using self-attention mechanisms and deep CNN and reinforcement learning as a way of interpreting the loan and trading decisions

Learn more about this project

Blockchain Technologies for Cooperation and Risk Reduction of Green Energy Suppliers

The project will develop a framework of multilateral cooperation between green energy suppliers and ancillary service providers, to reduce their overall risk profile and be more competitive in bidding into energy markets. This will be enabled through the use of blockchain technologies and evaluated on NY grid data

Learn more about this project

Dynamic Asset-Liability Management with Immunization and Risk Control

We intended to propose a notion of duration for a stochastic cash flow and to incorporate a duration constraint in a portfolio optimization problem, which is dedicated to matching (in a statistical sense) a random stream of liabilities. The question is particularly relevant when securities sensitive to changes in interest rates such as bonds are used but risky assets and other factors may be considered as well. The bonds may default within the given time horizon. We shall incorporate the statistical estimation of the duration into the portfolio optimization problem.

Learn more about this project

Knowledge Management Strategies in Financial Enterprises

Knowledge management strategy is critical to the success of financial enterprises, such as, asset management firms, Firms with organizational structures supporting coherent knowledge management strategies (talent recruitment, external/internal collaborations. thought leadership etc.) have a competitive advantage. Strategy design impacts the key performance indicators (KPIs) of an asset management firm.

Learn about this project

Defining and Defying Decentralized Finance (DEFI)

There are many problems with existing and emerging Decentralised Finance (DeFi) products due to smart contract bugs, oracle manipulation, and governance based majority attacks. This project aims to characterised DeFi products, mine transaction graphs, and analyze smart contract code to detect, prevent, and recover from such problems.

Learn more about this project

A Model-Free Quantum Simulation Framework for Volatility Derivatives Valuation

We investigate how quantum engineered system can be combined with a model-free simulation framework to utilize the plethora of available financial data to design a rapid quantum enhanced general simulation platform for the valuation and real-time risk analysis of financial options

Learn more about this project

Explainable ML for Credit Risk Analytics

This research project aims to assess the validity of explainable AI/ML(XAI) tools and models in the context of credit analytics by comparing the time-series and cross-sectional stability of XAI algorithms (LIME, Shapley values, etc. )with regard to several common ML algorithms (including logistic regression, tree-based models, and deep neural networks) on real consumer credit data.

Learn more about this project

AI-Driven Early Estimates of State for Predictive Robust Portfolio Allocations

State (market, sector, company) is stable as opposed to returns which are subject to volatility. We will build methods to predict state and robustly estimate state-dependent return distribution (HMM and deep network technologies). Our input features are flexible: financials technical indicators, macroeconomic indicators, etc. Given robust estimation of state, we use state to perform dynamic portfolio allocation.

Learn more about this project

Text and Knowledge Graph Mining for Asset Management

Effective text and knowledge graph mining, leveraging the vast troves of textual data spanning disclosures, reports and reviews to extract insights and features for asset management.

Learn about this project

Search form