Innovative Technology at the Interface of Finance and Operations: Volume I

Foreword
Contents
Contributors
1 Blockchain and Other Distributed Ledger Technologies, an Advanced Primer
1.1 What Is a Blockchain?
1.2 Blockchain Versus Other Forms of Distributed Ledger Technologies (DLTs)
1.3 The Bitcoin Example
1.4 Consensus Mechanisms
1.5 Performance and Scalability
1.6 Smart Contracts and Oracles
1.7 Tokens
1.8 Data Structure and Immutability
1.9 Anonymity and Pseudo-Anonymity
1.10 Security
1.11 Conclusion
References
2 Operational and Financial Implications of Transactionalizing Multi-Machine Maneuvers in Self-Organizing Autonomous Systems
2.1 Introduction
2.2 Negotiation over Shared Space
2.3 Priority and Trust in Collective Maneuver Planning
2.4 Machine Trust
2.5 An Earned Machine Trust Ecosystem
2.6 Collective Maneuver Planning: A Summary
2.7 Considerations Related to Value Distribution
2.8 Implications for Roadway System Operations and Finance
2.9 Conclusions
References
3 Interface of Operations and Finance: A Tutorial
3.1 Introduction
3.2 The Blind Men and a Firm Parable
3.3 A Brief Finance Primer for OM Researchers
3.3.1 Corporate Finance
3.3.1.1 MM World
3.3.1.2 Static Tradeoff Theory: Interest Tax Shield vs. Bankruptcy Costs
3.3.1.3 Moral Hazard as Financial Frictions
3.3.2 Capital Markets: A Few Key Results from the Asset Pricing Theory
3.3.2.1 The Fundamental Pricing Equation: Economic Motivation and Derivation
3.3.2.2 Risk-Return Trade-Off and Factor Pricing Equations: CAPM
3.4 A Brief OM Primer for Finance Researchers
3.4.1 Overview: A Struggle to Reconcile Demand and Supply
3.4.2 Order Commitment and the Newsvendor Model
3.4.3 Demand Risk (or Inventory) Pooling
3.4.4 Little Flexibility Goes a Long Way
3.4.5 Dynamic Management of Resources: Inventory and Capacity Expansion
3.4.5.1 Order-up-to Inventory Policies
3.4.5.2 Balancing Fixed and Variable Costs: (s,S) and EOQ Inventory Policies
3.4.6 Endogenous Sources of Variability in Supply Chain Procurement: Bullwhip Effect
References
4 The Past, Present, and Future of the Payment System as Trusted Broker and the Implications for Banking
Abbreviations
4.1 Introduction
4.1.1 Background and Context
4.1.2 Payment System Basics
4.1.2.1 Types of Payments
4.1.2.2 Definitions and Terms
4.1.2.3 System Architecture
4.1.3 Scope of this Paper
4.2 A Short History of Payments to the Present Day
4.2.1 Money and Payments in the Ancient World
4.2.2 Payment History from the Medicis to Modern Banks
4.2.3 Essential Linkages between Payment and Credit Services
4.2.4 Brief Overview of the Current Payment Architecture
4.2.4.1 Domestic
4.2.4.2 International/Cross-Border
4.2.5 Locus of Trust in the Current Payments System
4.2.6 Lessons and Opportunities for the Future
4.3 Trends Driving the Present Transformation of the Payment System
4.3.1 Changes in User Needs and Preferences
4.3.2 Technological Developments
4.3.3 Increasingly Cashless Societies
4.3.4 Peer-to-Peer (P2P) Payments
4.3.5 Growth in Value-Added Payment-Overlay Services
4.3.6 Instant Payments
4.3.7 Convergence between Instant Payments and Lending
4.3.8 Funding Needs and Liquidity Risks
4.3.9 Changes in Financial Intermediation
4.3.10 The Ongoing Impact of COVID-19
4.4 The Emergence of New Payment Platforms
4.4.1 Mobile Payments
4.4.2 Invisible Payments and IoT Technology
4.4.3 Digital Currencies
4.4.4 Distributed-Ledger Technology
4.4.5 Social Networks that Become Payment Networks
4.4.6 Major Banking and Technology Consortia
4.4.7 Why DLTs May Not Solve the Trust Problem
4.5 Developments to Watch
4.5.1 Standardization and Interoperability
4.5.2 The Changing Economics of Payments
4.5.3 Fintech Partnerships and Other Ecosystem Developments
4.5.4 Changes in Regulatory Climate and Philosophies
4.5.5 Central Bank Digital Currencies
4.5.6 Payment System Features
4.5.6.1 Real-Time Anti-Money Laundering (AML) and Fraud Control
4.5.6.2 The Erosion of User Dependence on Bank Security Credentials
4.5.6.3 Tokens and New Settlement Arrangements
4.5.6.4 The Persistence of Legacy Products and Systems
4.5.7 The Rise of New Bank Archetypes
4.5.8 Lessons from Emerging Markets for Developed Markets
4.6 Synthesis: The Changing Locus of Trust
4.6.1 Transformative Trends in Financial Disintermediation
4.6.2 The Competitive Threat from Large Digital Enterprises
4.6.3 The Importance (or Not) of Infrastructure Ownership
4.6.4 How Losing the Payment-Broker Role Could Disrupt Credit and Other Banking Services
4.6.5 The Changing Nature of High-Trust Customer Relationships
4.6.6 Plausible Scenarios for Future Disruptions and Transformations of Payment Systems
4.7 Conclusion
References
5 Machine Learning in Healthcare: Operational and Financial Impact
5.1 Introduction
5.1.1 Short, Simplified Overview of Hospital Operations
5.2 Matching Hospital Capacity with Demand
5.3 Readmission
5.4 New and Emerging Machine Learning and AI Technologies
5.4.1 Chatbots
5.4.2 Automation in Imaging
5.5 Fairness and Transparency
5.6 Conclusions
References
6 Digital Lean Operations: Smart Automation and Artificial Intelligence in Financial Services
6.1 Introduction
6.2 Diagnostic to Assess the Opportunities of Digital Operations
6.3 Digital Operations at Euroclear
6.4 The Promise of Smart Automation
6.5 Discussion
References
7 Applied Machine Learning in Operations Management
7.1 Introduction
7.2 A Brief History of ML
7.3 Supervised Learning
7.3.1 General Introduction to Supervised Learning
7.3.2 Supervised Learning for Descriptive Analysis in OM
7.3.2.1 Prediction with Supervised Learning
7.3.2.2 Causal Inference with Supervised Learning
7.3.3 Supervised Learning for Prescriptive Analysis
7.3.3.1 Prediction, Then Prescription
7.3.3.2 Better Prescriptiveness
7.4 Unsupervised Learning
7.4.1 General Introduction to Unsupervised Learning
7.4.2 Unsupervised Learning for Descriptive Analysis
7.4.2.1 Unsupervised Learning for Prediction
7.4.2.2 Using EM Algorithms in Choice Model Estimation
7.4.3 Unsupervised Learning for Prescriptive Analysis
7.5 Bandits and Reinforcement Learning
7.5.1 Multi-Armed Bandits
7.5.1.1 Popular Variants
7.5.1.2 Dynamic Pricing
7.5.2 Reinforcement Learning
7.6 Future Directions
References
8 Artificial Intelligence and Fraud Detection
8.1 Introduction and Motivation
8.2 Challenges of Fraud Detection
8.2.1 Problems with Fraud and Machine Learning in General
8.2.2 Problems that Are Specific to Accounting Fraud and Machine Learning
8.3 Practical Considerations in Model Building
8.3.1 Data
8.3.1.1 Data and Fraud
8.3.1.2 Data and Financial Statement Fraud
8.3.2 Methods
8.3.3 Evaluation Metrics
8.3.3.1 Evaluation Metrics in General
8.3.3.2 Evaluation Metrics for Accounting Fraud Prediction Models
8.3.4 Caveats about ML Models
8.3.5 Distinction between Prediction and Causal Inference
8.4 A Brief Overview of Existing Academic Research on Fraud Detection
8.4.1 Fraud Prediction Using Machine Learning in the Nonaccounting Academic Fields
8.4.2 Fraud Prediction in the Accounting Literature
8.4.3 Future Research Directions
8.5 Conclusion
References
9 AI in Financial Portfolio Management: Practical Considerations and Use Cases
9.1 Three Basic Goals for AI in Financial Portfolio Management
9.2 Using AI in Portfolio Management: How it Works
9.2.1 Introduction to the Various Types of AI Used in Financial Asset and Portfolio Management
9.2.2 Key Portfolio Optimization Concepts
9.2.3 Brief Reference to Optimization Regimes Enabled by AI
9.3 Investment Process Optimization
9.3.1 New Alpha-Generation Strategies for the Portfolio
9.3.2 The Human in the Machine
9.3.3 Research Applications
9.3.4 Trade Execution Improvements
9.4 Operational Efficiency
9.4.1 Front Office
9.4.2 Middle Office
9.4.3 Back Office
9.5 Customer Experience: Major Consumer Trends Driving the Adoption of AI
9.5.1 How the Shift to Digital and Mobile Intersects with AI to Create New Service Models
9.5.2 The Drive for Convenience
9.5.3 Personalization Enabled by AI Technology Resulting in Mass Customization
9.6 Business Outcomes
9.6.1 New Service Models
9.6.2 New Business Models and Offerings
9.6.3 New Investment Strategies
9.7 Further Considerations
9.7.1 Regulatory Compliance and Security
9.7.2 Algorithmic Bias
9.7.3 Transparency and Control
9.8 Conclusion
References
10 Using Machine Learning to Demystify Startups' Funding, Post-Money Valuation, and Success
10.1 Introduction
10.1.1 Financing
10.1.2 Post-Money Valuation
10.1.3 Success
10.2 Data
10.3 Methodology
10.3.1 Sectoral Clustering: Latent Dirichlet Allocation
10.3.2 Predicting Post-money Valuation: ElasticNet
10.3.3 Predicting Post-money Valuation: XGBoost
10.3.3.1 Hyperparameter Tuning for XGBoost Using Bayesian Optimization
10.3.4 Predicting Success: Neural Network
10.4 Results
10.4.1 Sectoral Clustering
10.4.2 Predicting Post-Money Valuation
10.4.3 Predicting Success
10.5 Recommendations
10.6 Conclusions, Limitations, and Future Work
References