Today's world is full of enormously inter-linked countries and businesses. Therefore, business planning decisions should take into account the linkages between different business segments as well as countries. Government policies also affect businesses not only in the home country but also elsewhere. However, typically business analysts who work on forecasts and planning do not account for changes in terms of policies and supply chains in their countries and elsewhere. In this session, we propose a framework that marries data-intensive computational methods employed in economic modeling with business decision making. The framework we propose is well-known in computational economics, called Computable General Equilibrium (CGE) model. It captures equilibrium in all markets in the economy, and tracks the movement from one equilibrium to another, in the presence of a perturbation from outside the system - like a policy shock, technological change or disaster. This framework employs publicly available rich information on linkages between countries and business segments to evaluate the market changes that can arise from major global and local policies as well as technological changes and disasters such as pandemics. This has immense potential to facilitate a scientific approach to handle future uncertainties involved in analytics for business decisions. This can be used for manufacturing sector or any other sector, to adjust their business strategies to external environments. Technically, we will discuss about data concepts in supply chain economics - called Input Output matrix-based models, which may further be built with behavioral equations that may be parametrized using AI/ML methods or conventional econometric techniques. In real time, we will show a classic example of this kind of a model - an open source global supply chain CGE model, named Global Trade Analysis Project (GTAP: We will conduct simulations to show how US-China trade war can affect different sectors in India, and how Covid-19 pandemic is impacting several sectors in India, based on our recent research with several international public organizations (ADB, UN, US Chamber of Commerce etc.) We shall provide study materials and links to download the software needed, ahead of the session. We will need atleast 45 minutes or 90 minutes (or even half a day session), to make sure that the audience can learn something actionable from this. These tools have been used by top consulting firms like McKinsey and almost all international organizations and governments, for example, for key decision making. There is also a lot of scope to use standard AI/ML tools herein. The purpose of this session is to both introduce these tools to the audience and also encourage them to collaborate on developing AI/ML solutions for some challenges in these tools, such as lack of accuracy in capturing the behavior of different components of the economy and forecasting the future.


Outline/Structure of the Demonstration

1. Motivate the topic and explain what is global supply chain modeling, GTAP (Global Trade Analysis Project, Purdue University) in particular, and why is it relevant for decision making, with some example applications on the slides. (10 minutes)

2. Explain the thoeretical features of global supply chain models (10 minutes)

3. Show the application on the software (RunGTAP) of US China Trade Wars and Covid-19 economic impact. (15 minutes)

4. Explain the scope for AI/ML in these models (5 minutes)

5 minutes for Q& A if we get a 45 minutes session.

If we do a 90 minutes session:

We will still do 1 and 2 at the same pace as above (20 minutes in total); then we will do the following:

3. Data construction for global supply chain models: different data sources (Input Output tables, macroeconomic data, trade, etc), methods of reconciliation (RAS, entropy, etc), with some demo with the code. 15 minutes

4. Modification of GTAP data to a desired level of aggregation of sectors and regions, and some methods to change the data (Altertax, GtapAdjust, etc.) , with extensive demo on how to do these things. 15 minutes

5. Show the application on the software (RunGTAP) of US China Trade Wars and Covid-19 economic impact. (20 minutes) : extensive demo, with complete explanation of all inputs into the model, and discussion of results from the model

6. Explain the scope for AI/ML in these models (10 minutes)

10 minutes for Q&A

If we do only 20 minutes, we can only cover the first two topics in 10 minutes, then do a quick demo in 8 minutes and keep 2 minutes for Q&A. This may not be the most effective.

Learning Outcome

If we do the 20 minutes session: The audience may learn some terminology involved in global supply chain economic modeling, GTAP, etc., and be able to have some basic understanding of such models and their applications.

If we do 45 minutes session: In addition to the above, they can also run a basic version of off-the-shelf GTAP data and model on their own, and help support decisions for business strategy and public policy, with some practice and experience.

If we do 90 minutes session, they may be able to critically appreciate the data and model, and modify the datasets to reflect the realities even more, and understand the results in a rigorous and practical way, for decision making. They will be able to choose between different assumptions to effectively predict and forecast impact of different policies/technologies/disasters on global supply chains and specific industries.

Target Audience

Executives involved in decision making at various levels, business analysts, data scientists supporting strategic decisions, policy analysts, management consultants, economists and data scientists at large

Prerequisites for Attendees

No prerequisites, but a very basic, high-level knowledge of economic theories may help to some extent ; a glimpse of this wiki page for example:

We would also recommend installing the following software, and if possible go through some of the materials and videos in the following links: (may need to sign up/register in GTAP website)



schedule Submitted 2 years ago

  • Kavita Dwivedi

    Kavita Dwivedi - Portfolio Valuation for a Retail Bank using Monte Carlo Simulation and Forecasting for Risk Measurement

    Kavita Dwivedi
    Kavita Dwivedi
    Head Data Science
    schedule 2 years ago
    Sold Out!
    20 Mins

    Banks today need to have a very good assessment of their portfolio value at any point in time . This is both a regulatory requirement and an operational metrics which helps banks to assess risk of their portfolio and also calculate the Capital Adequacy that they need to maintain at portfolio levels , product levels and all of these aggregated at Bank level.

    This presentation will walk you through a case study which will discuss in detail how we went about calculating Portfolio value for a Home loan on a sample data . The bank wanted a scientific /statistical approach to this as they could take this to regulators for approval and thus convince them about the capital that they have for a particular portfolio.

    The other interesting dimension was that in case the bank wants to sell a particular loan book to another bank /third party financial institutions they would be able to quote a price within the confidence interval of the calculated price. The same model/tool could be also shared with the buyer to convince them on quoted price and will make the negotiation and selling smooth.

    We have used Monte Carlo Simulation on historical data of the portfolio to measure the Portfolio Value for the next 5 years of a Home loan Portfolio. It is a two step modeling process with Machine Learning Models to predict default and then further using simulation to calculate Portfolio value year on year for next 5 yrs taking in account diminishing returns too.

    The presentation will take you through the approach and modeling process and how Monte Carlo Simulation helped us deliver the same to Customer with high accuracy and confidence level.

    This is a real case study and will focus on why Risk Measurement is important and why Basel , CCAR implementation across banks worldwide helps the Central Banks to manage risks in case of a financial downturn or Black Swan events.