QUALITATIVE ANALYSIS OF BUSINESS DYNAMICS MADE EASY BY VENSIM

System Dynamics Modeling is powerful tool to find about system's non linear behaviour over time using stock and flow, internal feedback loops and time delays. It is a computer-aided approach to policy analysis and design. It applies to dynamic problems arising in complex social, managerial, economic, or ecological systems, literally any dynamic systems characterized by interdependence, mutual interaction, information feedback, and circular causality. It qualitatively highlights multiple aspects and behaviour of system which may be ignored or unnoticed in planning and designing. It allows to see the picture of future for such system designs.

We would like our readers to become familiar with this powerful tool of qualitative analysis by introducing example scenarios. European ATM system will be analysed and using features of Vensim we will show how can we cater the vulnerabilities in both systems. These are just examples, the software can be utilized in numerous possible fields.

European ATM case will be analysed using system dynamics modeling, finding the relationship and behavior of multiple factors or variable involved in the design. First the system model will be designed in accordance with the given conditions and then system would be simulated to show the resutls with different inputs and variations in the variables and factors. The stock and flow, time delay and feedback loop for problem and solution of complex system would be simulated using Vensim simulation tool. The resultant graphical information for each simulation will be attached to indicate the behavior of system as per requirement. 

Modeling System Dynamics for European Case:

We will develop a problem archetype showing that when the burden of proof is on customers the intention to contain fraudulent transactions to exert pressure on the customers will cause bank complacency, negligence of security in the first place and ultimately more fraudulent transactions as the security vulnerabilities are exploited by malicious agents.

Now we identify the problem archetype as one of the four types (underachievement, relative achievement, out of control, relative control).

Given that the European banks had opted for the model of putting the burden of proof on the customers, suggest a solution link to fix the vulnerabilities in the ATMs.

Figure attached here indicate the Problem and Solution archetypes for above mentioned problem.

For previously described European ATM case we rely on intended consequence (IC) and the unintended consequence (UC) – in other words, it implements the full problem archetype. It is required to simulate the models so as to learn how the key variables behave over time and to inspect the models’ variables and their relations in order to understand and explain the structure of the models. Finally to discuss critically the models’ assumptions. Then we enhance the system dynamics model of the problem archetype by implementing a solution loop. In other words,we are asked to implement a system dynamics model of the solution archetype.

European ATM Case (IC) Model:

Fig. 3 shows the system dynamics model that expresses the intended consequence of the bank’s action, i.e. putting the burden of proof on the customer if (s)he claims that a fraudulent transaction has happened. Since the model ignores the unintended consequences, you can think of it as what would happen in an ideal world where the only possible fraud would occur from some bank customers. In other words, one ignores that there may be professional crooks ready to exploit the vulnerabilities of the ATMs. We are required to perform following tasks as a solution to the given problem:

1. to simulate the models so as to learn how the key variables behave over time; 

The simulation requires to generate a feedback link from the flow ‘’cumulated frauds’’ to the stock ‘’customer’s fraud rate’’. The simulation are without any professional crooks involvement which will be done in next simulation. Setting the parameters or key variables with following values, following graph dispalys the final result.

• Initial fraud rate: 15 Frauds per month
• Bank’s perception time: 1 Month
• Bank’s acceptable fraud rate: 1 Fraud per month

The effect of change in above mentioned variables is positively corelated. Increase in values of these factors result in rise of vertical scale. Bank’s acceptable fraud rate is less effective in this regards as compared to other two. The graph clearly indicates that the cumulated frauds are increasing with the passage of time, hence within 2 years time they reach upto 8000 almost.

2. to inspect the models’ variables and their relations  in order to understand the model;

Name	Type	Equation	Unit
Initial customer’s fraud rate	Constant	10	Fraud per month
Customer’s fraud rate	Auxiliary	Initial fraud rate + (Cumulated frauds/Rate of change of customer’s fraud rate)	Fraud per month
Cumulated frauds	Level	Customer’s fraud rate	Fraud
Perceived customer’s fraud rate	Auxiliary	(Initial Customer’s fraud rate + Customer’s fraud rate)/Bank’s perception time	Fraud per month
Bank’s perception time	Constant	1	Month
Fractional deviation from acceptable fraud rate	Auxiliary	Perceived Cutomer’s fraud rate – Bank’s acceptable fraud rate	Fraud per month
Bank’s acceptable fraud rate	Constant	1	Fraud per month
Time to change rate of change	Auxiliary	Fractional deviation from acceptable fraud rate	Month
Effect of burden of proof on cutomers	Auxiliary	Rate of change of customer’s fraud rate/Time to change rate of change	Dmnl
Rate of change for customer’s fraud rate	Level	5	Fraud per month

3. to explain the structure of the models; 

Variable	Variable explanation
bank's acceptable fraud rate	The bank accepts this low fraud rate as "acceptable" in the sense that enforcing a lower fraud rate would be too costly (the return on investment would be too low). Consider this as the bank's residual risk.
cumulated frauds	Cumulated frauds are result of initial customer’s fraud over time.
customers' fraud rate	No of frauds per month
effect of burden of proof on customer	It is the stock for rate of change of customer’s fraud rate.
fractional deviation from acceptable fraud rate	Depends on percieved customers fraud rate and banks acceptable fraud rate
initial customers' fraud rate	Initially assumed frauds per month
rate of change for customers' fraud rate	Change in no of fruads occuring in a month
time to chng rate of change	Rate of change in frauds is defined while it is the time to change it.
Perceived customer’s fraud rate	How much fruads encountered in a month as percieved by the customer
Bank’s perception time	It is the time defined for percieving frauds in months

4. to discuss critically the models’ assumptions;

The model is designed using Vensim simulation tool to study the system dynamics and non linear behaviour over time for the given problem. European ATM problem is thus analysed deeply to see the future behaviour based on proposed solution. This model depicts the problem archetype Intended Consequence. Multiple factors playing role in the ATM system are indicated by arrows to show their internal relation. Beauty of Vensim design is that it delivers beyond the hypothetical approach. The variables or the real life factors involved in are connected with mathematical equations and thus the quantitative results for different inputs is achievable while we simulate the design. This is one particular problem archetype design that shows the result in form of graphical information. There could be several designs other than this for the same problem and also solution designs. Factors involved in the whole model could also be different but it is one approach to analyze the given scenario and a way to move for the sultion based on qualitative analysis and quantiatively assumed values.

European ATM Case (IC-UC) Case:

Fig.5 shows the system dynamics model that expresses the intended consequence of the bank’s action, i.e. putting the burden of proof on the customer if (s)he claims that a fraudulent transaction has happened, as well as the unintended consequence of launching a fraud epidemic in that professional crooks exploit the vulnerabilities of the ATMs. The crooks thrive, since the bank’s antifraud measures will be directed towards the bank’s customers (who have the quite impossible task to prove that they did not commit the fraud themselves). We are required to perform the following tasts, as a solution to the given problem:

1. to simulate the models so as to learn how the key variables behave over time; 

The simulation performed above is now added with professional crooks involvement to make it more real life scenario. It is now more vulnerable to threats. Key variable are slightly changed from above simulation plus the following additional key variable vlaues are set;

Crook’s perception time of ATM security: 1 Month

Bank’s acceptable fraud rate: 1 Fraud per month

Unit factor: 1 Dmnl

Bank’s perception time: 1 Month

Initial fraud rate: 10 Fraud per month

With these parameters set, are variable in simulation to find effect on resultant graph, the following graph is found;

Fig 6 shows the graph that depicts the comparison analysis of ‘’customer’s fraud’’ and ‘’ crook’s fraud’’. Both show almost similar result. . The graph clearly indicates that the cumulated frauds are increasing with the passage of time, hence within 2 years time they reach upto 6000 almost.

2. to inspect the models’ variables and their relations  in order to understand the model;

Undermentioned variables and equations are additional to the previous simulation. The ATM security factors and crook’s involvement has resulted in these additional variables and equations to final simulation.

Name	Type	Equation	Unit
Crook’s fraud rate	Auxiliary	Crook’s perception time of ATM security – ATM security	Fraud per month
Crook’s perception time of ATM security	Constant	1	Month
ATM security	Level	5	Fraud
ATM security obsolescence	Auxiliary	ATM security – Security obsolescence time	Dmnl
Security obsolescense time	Auxiliary	Bank’s complacency	Dmnl
Bank’s complacency	Auxiliary	Effect of burden of proof on customer/Unit factor	Month
Unti Factor	Constant	1	Dmnl per month

3. to explain the structure of the models; 

Variable	Variable explanation
ATM security	ATM Security is the factor to mitigate the threats in terms of customer’s fraud and crook’s frauds.
ATM security obsolescence	ATM security no longer wanted, even though it is still in good working order.
bank's complacency	Bank’s self satisfaction about it being secure. This factor results in vulnerabilities and hence more chances of frauds.
crooks' fraud rate	Crook’s fraus occuring in a unit time, set as month.
Crooks' perception time of ATM security	This is how a crook percieves about ATM security in terms of perception time whereas perception time is some one’s own perception of duration of indefinite and continuous unfolding of events.
unit factor	This factor, with the value 1 Dmnl/Month, is used the make the variable "bank's complacency" dimensionless
Security obsolescence time	Time for security to becom obsolescence

4. to discuss critically the models’ assumptions.

In the previous model, customer’s involvement in fraudulant transactions for European ATM case was highlighted. While in this model we study the crook’s involvement in fraudulant transactions as well and see the behavour of graph on cumulated frauds variable.This include unintended consequnce. Crook’s fraud is taking effect on the basis of vulnerabilities in the form of security obsolescense, bank’s complacency and ATM security obsolescence. Thus the cumulated fraud shown in graph has two traces by including customer and crook. Multiple key variables can be changed while running the simulation to see the effect on model. A general qualitative analysis would result in favour of designed model. In the following section we would work for solution loop and suggest the solution archetype simulation model.

Extending Simulation Model:

So far the problem archetype are designed on the basis of System Modeling using Vensim to find out the results of our assumptions based on different factors in the form of graphical results. Now it is time to design the solution archetype or solution loop that will deliver the final solution to the problem under discussion. This will let us know how the customer’s frauds and crook’s frauds can be dealt. Although it is clear from previous drawings and simulations that the ATM security need to be improved but the final solution model is as follows:

As shown in fig.7, the solution loop is between ‘’cumulated frauds’’ and ‘’ATM security’’ whereas ‘’cumulated frauds’’ are combined consequence of both customer’s fraud and crook’s fraud. This solution loop emphasize on the variable ‘’Bank’s spending to fix vulnerabilities’’. Greater this factor would result in improved ATM security and hence reduce the threats of all sorts of frauds.

It required to modify ‘’ATM security’’ flow with following equation to take the effect;

ATM security = Bank’s spending to fix vulnurabilities x (ATM security – ATM security obsolescence)

Solution Graph:

System modeling helps in analyzing future statistics of any given scenarion in advance. Fig.8 shows the resulted graph of final solution model simulation for European ATM case. It is combined result of intended consequence, unintended consequence and solution case. The blue coloured graph line is showing the effect of ‘’Bank’s spending for vulnerabilities’’. Customer’s frauds and crook’s frauds in the system are inceasing with passage of time highlighted with red and green lines respectively. The blue line proves our solution loop has the effect to reduce fradulant transaction by introducing Bank’s attention in the form of Bank’s spending for vulnerabilities by improving ATM security.
                                                    ----------------------------------------------------------------

''Copy the following link to your browser's URL to download Vensim for your system.''

Vensim Download.

Watch the video for details about Vensim, how does it work?