Demand Forecasting for a Bank

Client Issue

One of the top Russian banks was interested in enhancement of the assets and liabilities management. One way to achieve the goal was to build multivariate model for estimation and forecasting full range of the bank’s products sales and simulating sales under different scenarios prepared either by treasury department or by commercial department.


The forecasting and modelling results are used for a wide range of tasks:

  1. Risk management strategy
  2. Business planning
  3. Feasibility of changes in products interest rates assessment
  4. Marketing campaigns effectiveness assessment

Concerned departments for this project is Marketing department, Commercial department, Treasury department.

DAI Solution

DAI developed system that utilizes both wide domain banking expertise and machine learning algorithms that took into account data on product sales, marketing expenses, media ratings (both for the client and its competitors) and macroeconomic. As a result, domain expertise was proven and complemented by machine learning algorithms.


DAI applied a complex approach on achievement of customer’s goals including the following steps:

  1. Problem framing.

    Deloitte collected expectations, issues and requirements from business departments and concerned employees. DAI specialists conducted interviews with stakeholders (treasury, marketing, business, IT), clarified requirements for the system and managed client’s vision
  2. Data collection and preprocessing.

    On this stage in addition to database with world and Russian macroeconomics indicators gathering by Deloitte, the bank provided huge amount of data from different sources for subsequent data preprocessing and analysis. Examples of data used includes:
    • Product sales
    • Accounting information for each product
    • Product conditions
    • Marketing budget
    • Media ratings
    • Macroeconomic drivers (i.e., currency rates, industrial indexes, price indexes etc.)

    This step included data consistency check between different sources and inconsistencies treatment, handling with duplicates and null values, dictionary forming and transformation, generation of new features, statistics calculation and reporting both detected problems to client and solution for them.

  3. Data processing and algorithms engineering.

    This step DAI started with correlation and distribution analysis of used factors in order to notice top-level dependencies (both linear and non-linear) and what’s more important to notice change of different factors influence. Taking it into account mathematical model was built for each type of the product. The model successively incorporated three base models with different number of factors to achieve the best forecast.

    Also in accordance with the client’s requests DAI researched model’s sensitivity and created tool to model different sales outcomes under different conditions.Achieved result was proven in future 6 month observations and demonstrated the highest quality among other systems tested by the client
  4. User interface prototyping.

    Together with the models development DAI UI/UX group prototyped user interface the solution based on interviews with stakeholders and conducted surveys. The interface allowed users from concerned departments to work with the models, create scenarios, analyze different types of products and create reports based on results

Methods

  • Machine Learning
  • Time Series Analysis
  • Econometrics
  • Interface prototyping

Outcomes

Completed system built at the junction of domain banking expertise and machine learning algorithms that uses data on product sales, marketing expenses, media ratings (both the client’s and its competitors’) and macroeconomic.

 

The system allows:

  1. Analysis elasticity of portfolio volume in dependence on products parameters
  2. To provide forecasting of portfolio volume for the selected time horizon
  3. To research different outcomes under different conditions and scenarios set by the user

How our solution affects the Shareholder Value

Shareholder Value

  1. Revenue Growth
    1. Volume
      1. Acquire New Customers
      2. Retain & Grow Current Customers
      3. Leverage Income Generating Assets
    2. Price Realization
      1. Strengthen Pricing
  2. Operating Margin (after taxes)
    1. Selling General & Administrative (SG&A)
      1. Improve Customer Interaction Efficiency
      2. Improve Corporate/ Shared Service Efficiency
    2. Cost of Goods Sold (COGS)
      1. Improve Development & Production Efficiency
      2. Improve Logistics & Service Provision Efficiency
    3. Income Taxes
      1. Improve Income Tax Efficiency
  3. Asset Efficiency
    1. Property, Plant & Equipment (PP&E)
      1. Improve PP&E Efficiency
    2. Inventory
      1. Improve Inventory Efficiency
    3. Receivables & Payables
      1. Improve Receivalables & Payables Efficiency
  4. Expectations
    1. Company Strenghs
      1. Improve Managerial & Governance Effectiveness
      2. Improve Execution Capabilities
    2. External Factors

How our solution affects the Shareholder

Revenue Growth – Optimizing Existing Products Parameters

Operating Margin – Improve Operation Efficiency

Business Management – Improve Efficiency in Interaction with Customers