University:
Massachusetts Institute of Technology
Course:
15.501/516 | Accounting
Academic year:

2023
Views:

443

Pages:

19
Author:

Rebecca Chaney

Applications of System Dynamics Agenda • How can you help a client • The Big Enchilada • The Standard Method How Can You Help a Client? (Objectives of An SD Effort) • Point prediction • Managing better Point prediction: Examples • Litigation (Retrodiction) • Commodities Markets • Contract bid Point Prediction: Problems • • • Difficult Risky: Benefit comes only at the end Less useful than commonly believed: “If I only knew what was going to happen, managing would be simple” – Simple cases: Litigation, financial speculation – Tough cases: You want to change the future Managing Better: Examples • Overtime policy • Pricing policy • Capacity expansion Managing Better: Problems • Hard to sell – Clients often pose their problem as one of prediction – Difficult to explain what a policy is in the abstract • Difficult to appreciate what you’ve learned – “I knew it all along” • Often no clean-cut finish to the project Difficult to recall what you got out of the process Recommendation Phase 1 Phase 2 Impact Outsource Receivables X Use Hedging & Futures markets X Tighten Customer payment & credit policies X Change or abandon CFCT measurement X More reliable demand forecasting X Reduce target inventories & accept increased risk X X X + Reduce Time to ramp up flywheel sales X + Faster planning cycle X Improved dialogue & communication with all stakeholders X Improved (graphical) interfaces for managing planning cycle X Incorporate dynamic models in planning practices X Improve customer service metrics & feedback + + Why Its Tough to Realize What You’ve Learned Before After The Big Enchilada • “Small” Policy Model – Understand dynamics of issue – Create and explore policies • “Big” calibrated model – – – More precisely, when should we do X More precisely, how much should we do X More precisely, what is the benefit of X System Dynamics Calibrating Insight The Big Enchilada Calibrating Tuning Prediction Data Calibrating Calibrating Insight The Big Enchilada Insight: Modeling Tuning Prediction Data Calibrating Calibrating Insight The Big Enchilada Modeling Analysis Insight Insight: Analysis Tuning Prediction Data Calibrating Calibrating Insight The Big Enchilada Analysis Modeling Insight Insight: Modeling breakdown Tuning Prediction Data Calibrating Calibrating Insight The Big Enchilada Analysis Modeling Insight Modeling Info Gather Loops & Models Modeling Danger of big enchilada: Inadequate time and resources • Fast modeling • Skip model analysis • Skip data examination Concentrate on getting a model that fits the data The Standard Method • Reference modes Causal loop model • Simulation Model Analyzing model Creating and exploring Insights and Policies The Standard Method (detail) • 1) – – – • • • • • • • 2) 3) 4) 5) 6) 7) 8) Problem definition a) b) c) List of variables Reference modes Problem statement Momentum policies Dynamic hypotheses Model first loop Analyze first loop Model second loop Analyze second loop Etc. Insights and Policies Guiding the Initial Focus • Important to team members (so you get their time) • Dynamic (i.e. reference mode) • Enough time to work it without panic, not so much time that there is no conclusion