## John Sterman's Business Dynamics Book's Examples - AnyLogic Models

### Appendix A

Inv-WF with Noise

The inventory-workforce model developed in Chapter 19 (Widgets with Labor) augmented to include random variations in productivity. The random shocks in productivity can be switched between guassian (white) noise and pink (first-order autocorrelated) noise.

Pink Noise Normal

Structure to generate pink noise (first-order autocorrelated noise) from a normally distributed white noise input.

Pink Noise

Structure to generate pink noise (first-order autocorrelated noise) from a uniformly distributed white noise input. The pink noise output is asymptotically normal.

### Appendix B

Population Model

A simple linear population model used to illustrate techniques for numerical integration such as the Euler and Runge-Kutta methods.

### Chapter 4 — Structure and Behavior of Dynamic Systems

Pop and Carrying Capacity

A simple model of a population growing in an environment with a given carrying capacity. The carrying capacity can be constant or consumed by the population; nonrenewable or renewable. With different parameters, the model can generate the most important modes of behavior observed in dynamic systems and discussed in chapter 4, including exponential growth, exponential decay, S-shaped growth, S-shaped growth with overshoot and oscillation, and overshoot and collapse.

### Chapter 8 — Closing the Loop: Dynamics of Simple Structures

First Order Neg FB

The first-order linear negative feedback system, illustrating exponential decay. See section 8.3.

First Order Neg with Goal

The first-order linear negative feedback system with an explicit goal, illustrating exponential approach to a goal. See section 8.3.

First Order Pos FB

The first-order linear positive feedback system, illustrating exponential growth. See section 8.2.

Linear Population

First-order population growth model with linear birth and death rates. See section 8.4.

Nonlinear Population

First-Order population growth model with constant carrying capacity, illustrating S-shaped growth. See section 8.5.

### Chapter 9 — S-Shaped Growth: Epidemics, Innovation Diffusion, and the Growth of New Products

Bass Model

The Bass innovation diffusion model (innovation diffusion as driven by advertising and word of mouth feedbacks). See section 9.3.3.

Bass Repeat Purch Flow

The Bass diffusion model expanded to include repeat purchases of the product by adopters. See section 9.3.6; figure 9-22.

Bass with Discards

The Bass diffusion model expanded to include product discards and replacement purchases. See section 9.3.6; figure 9-20.

Logistic Model

The logistic growth model of population growth or innovation diffusion. See section 9.1.1.

Richards Model

The Richards population growth model. See section 9.1.3.

SI Model

The SI (Susceptible-Infectious) model of infectious disease. See section 9.2.1.

SI Innovation Model

SI epidemic model applied to innovation diffusion. See section 9.3.

SIR Model Threshold

The SIR epidemic model configured to illustrate the concept of the tipping point. See section 9.2.5.

SIR Model

The SIR (Susceptible-Infectious-Removed) model of infectious disease. See section 9.2.2.

### Chapter 10 — Path Dependence and Positive Feedback

Network Effect

A model of two firms competing for market share in the presence of network externalities, illustrating path dependence. The attractiveness of each firm’s product to customers depends on the size of the installed base (the network of users). See section 10.8.

Nonlinear Polya Process

The Polya urn model with a nonlinear function for the probability of selecting the color of the next stone to be added. The model illustrates path dependence. See section 10.2.1.

Polya Process

The classic Polya urn model in which the probability of adding a stone of a given color to a bag of stones equals the proportion of stones of that color already in the bag. Illustrates path dependence. See section 10.2.

### Chapter 11 — Delays

Adaptive Exp Random Walk

Adaptive expectations (first-order exponential smoothing) driven by a random walk. See section 11.3.1.

Adaptive Expectations

Adaptive expectations (first-order exponential smoothing) driven by a stationary random input. See section 11.3.1.

Nonlinear Smoothing

First-order exponential smoothing with a nonlinear time constant, allowing a different response time for increases and decreases. See section 11.4.1.

### Chapter 12 — Coflows and Aging Chains

Capital Labor Coflow

Simple model of a capital stock with a coflow to track the embedded labor requirements of each unit of capital. See section 12.2.

Capital Vintaging Coflow

Model of the capital stock for a firm, industry, or economy, disaggregated into an aging chain with different capital vintages, and including a coflow to track the factor input requirements embedded in each unit of capital. See section 12.2.2.

Faculty Aging Chain

Illustrates the dynamics of aging chains with the example of a university, tracking the hiring, promotion, and termination of faculty from assistant to associate to full professor. See section 12.1.6.

Hiring Chain 1

Simple model of a firm’s hiring and training process distinguishing between new, inexperienced employees and experienced employees; shows how experience and productivity vary with attrition rates, learning times, and growth. See section 12.1.7.

Labor Learning Curve

Models a labor force and their on-the-job learning from experience. Illustrates coflows with nonconserved flows. See section 12.2.1.

### Chapter 13 — Modeling Decision Making

Floating Goals

Models a negative feedback system in which the goal of the system is variable and adjusts to the past performance of the system itself. Illustrates floating, or eroding, goals. See section 13.2.10.

Hillclimb

A structure used to model local search by hill-climbing. The state of the system adjusts to a desired state which in turn is based on the current state modified by various pressures that indicate the gradient (the direction leading, at least locally, to higher performance). Useful in a wide range of models of local optimization, search, and learning. See section 13.2.12.

Price Discovery

Applies the hill-climbing structure to model the process of price discovery in a market where the market makers do not know the demand and supply curves of market participants. Provides a basic model of disequilibrium price adjustment. See section 13.2.12.

### Capter 15 — Modeling Human Behavior: Bounded Rationality or Rational Expectations?

Market Growth 1

The version of Forrester’s Market Growth model developed in chapter 15 (a model of a high-tech growth firm).

### Chapter 16 — Forecasts and Fudge Factors: Modeling Expectation Formation

TREND

The TREND function developed in chapter 16. Models the formation of growth expectations from the past history of an input time series.

### Chapter 17 — Supply Chains and the Origin of Oscillations

Stock Management 1

The stock management structure for controlling a stock in the presence of losses or usage and an acquisition delay for new units. In this variant, the desired supply line is based on the desired acquisition rate. See section 17.3.

Stock Management 2

The stock management structure for controlling a stock in the presence of losses or usage and an acquisition delay for new units. In this variant, the desired supply line is based on the expected loss rate. See section 17.3.

Stock Management 1st Order

The stock management structure for the first-order case in which there is no supply line of unfilled orders. See section 17.2.1.

### Chapter 18 — The Manufacturing Supply Chain

Multiplier Simul Eqns

Simple macroeconomic model based on the Keynesian consumption multiplier. Illustrates simultaneous initial value equations. See the Challenge on Simultaneous Initial Conditions in section 18.1.5.

W2Stage w DD FB

Expands the Widgets model with material inventories and order backlogs to represent two partners in a supply chain, one representing an OEM and one representing the supplier to the OEM. The delivery delay for the supplier is potentially variable, and the OEM responds by varying the desired supply line of materials on order. See section 18.2.

Widgets

A model of a manufacturing firm representing the supply line of production. Represents finished goods and work in process inventories, and the decision rules used to manage them in the face of unpredictable orders. See section 18.1.

Widgets w Mat Inv

Expands the Widgets model with order backlogs to include a stock of raw materials. See section 18.1.9.

Widgets w Backlog

Expands the Widgets model to include a backlog of unfilled orders. See section 18.1.7.

### Chapter 19 — The Labor Supply Chain and the Origin of Business Cycles

Labor w Layoffs

Applies the stock management structure to the labor supply chain for a firm, representing vacancies, vacancy creation and vacancy fulfillment along with the labor force, hiring, quits and layoffs. See section 19.1.

Widgets w Labor and OT

Adds overtime/undertime to the Widgets model with labor. See section 19.2.4.

Widgets w Labor

Integrates the Widgets model developed in Chapter 18 with the labor supply chain model Labor w Layoffs. See section 19.2.

### Chapter 20 — The Invisible Hand Sometimes Shakes: Commodity Cycles

Commodity1

The commodity industry model developed in chapter 20.

Price Sector

The price setting subsystem for the commodity model. Represents how market makers set prices through consideration of the demand/supply balance and expectations about the underlying value (expected equilibrium price) of the commodity. See section 20.2.6.

### Chapter 21 — Truth and Beauty: Validation and Model Testing

Inv-WF Noise Switch

The inventory-workforce model developed in Chapter 19 (Widgets with Labor) augmented to include random variations in productivity. The random shocks in productivity can be switched between two different sequences drawn from the same distribution to illustrate the difficulty of predicting the exact future values of a stochastic dynamic system even when it is perfectly specified. See section 21.4.7.

Summary Statistics

A module that computes summary statistics characterizing the historical fit of a model to data. Computes mean absolute percent error, (root) mean square error, R^{2}, the Theil inequality statistics, and others. See section 21.4.7.

*
*

*John Sterman, “Business Dynamics: Systems Thinking and Modeling for a Complex World”*

Business Dynamics
| |
---|---|

Author | John Sterman |

Publisher | McGraw Hill |

Publication date | 2000 |

ISBN | 0-07-231135-5 |

The book introduces systems dynamics modeling for the analysis of policy and strategy, with an emphasis on business and public policy applications. System dynamics is both a conceptual tool and a powerful modeling method. This allows the building of computer simulations of complex systems. These simulations can then be used to test the effectiveness of different policies on business outcomes.

The book includes CD with example models. You can find these models below. Just click on the model make and watch AnyLogic applet!