A majority of Indians have welcomed the Prime Minister’s bold initiative to flush out black money from the economy and polity of India. However, its implementation has been, to say the least, shoddy and technologically primitive. Flushing out high value old notes with new notes is primarily a supply chain management problem. With access to right data and analytics the managers of this supply chain, namely RBI and banks could have identified the most vulnerable points in terms of demand for cash and the optimum way to get new notes to these points. What follows is a brief analysis of the issues involved and how data analytics could have saved the current currency chaos.
Big Data and Big Data analytics
Big Data Analytics has two pillars- Big Data and Advanced Analytics. Big data, as its name suggests, is big in volume. Besides the size of data, big data has two more components- speed of data creation or velocity and variety of data to be handled or structure.
One of the major application area of BDA is Supply Chain Management. Supply chain analytics is the process by which the all stakeholders in the network can leverage supply chain information through the ability to measure, monitor, forecast, and manage supply chain business process. Supply of hard domestic currency to various economic agencies of a country is a gigantic supply chain management process in a country like India.
What is meant by currency management?
Currency management refers to the entire lifecycle of production of notes and coins, its delivery to currency vaults/chests in RBI or banks, its release through supply points, its receipt back to RBI / banks and reissue them if found suitable or destroy them otherwise.
How big and complex is India’s currency chain management process?
Currency management infrastructure consists of a network of 19 issue offices, 4,132 currency chests (including sub-treasury offices and a currency chest of the Reserve Bank at Kochi) and 3,813 small coin depots of commercial, cooperative and regional rural banks spread across the country. India today has around 80 thousand Scheduled commercial bank branches and 2 lac ATMs. All branches must receive notes and coins from their nearest currency chests small coin and depots.
A comparative picture of size of India’s currency operations is given below:
Challenges faced by currency management process owners:
- Capturing the structural and cyclical demands for currency and projecting demand with precision at all India level as well as state / district level
- Maintaining adequate supply of all denominations
- To achieve operational efficiency in distribution of notes by controlling cost of handling, distribution and security in transit/storage
- Ensuring durability of banknotes
The enormity of the process can be gauged from the volume of operations that take place at the RBI level only.
In 2015-16 accounting year of RBI, the cost of printing and remittance of currency incurred by RBI was 35.30 billion of rupees. The total cost of maintaining the entire infrastructure of currency management would be much higher.
How Supply chain analytics of big data can help to increase operating efficiency of India’s currency management process:
- Forecasting of Denomination wise currency demand:- Today RBI carries out denomination wise forecasting demand at macro level using estimated economic relationship wise total currency demand in value and then grafting on it denomination wise distribution using various time series techniques. BDA can help to create a better predictive model by using bottom up approach and top down approach sequentially and arrive at an optimal solution iteratively
- Estimating demand at a granular level:– By analysing data of cash outflow from individual ATMs/ branches and inflow to branches it should be possible to use machine learning algorithms to create a much better predictive model of demand at a granular level
- Optimization of transportation cost: Given demand data and supply capacity data it should be possible to create a gigantic multi-level transportation model to generate optimum supply schedule.
How Big Data could help avoiding currency chaos?
Once we have results of the BDA as stated above, we could have easily done the following:
- Identify the demand locations having relatively highest demand of high value notes as compared to the ones with maximum demand for lower denomination notes. Run your optimization models with new constraints and identify how the cash could be injected into the economy in in a calibrated manner. In the absence of such a model we have resorted to one size fits all. For example, in a Pedder Road branch we could have given more of 2000 notes while give only 100 rupees note in a branch located in a large slum.
- Identify locations like APMC market etc. supply them with only 500 notes to facilitate trading. Since we wanted a general solution we had to await supply of large quantity of 500 notes to distribute to all locations somewhat uniformly.
Thus using big data and BDA we could have substantially mitigated the pain of ordinary citizen and avoid the latest debilitating currency chaos.
Author- Dr Ashok K Nag is a former senior executive of Reserve Bank of India with more than 35 years of experience in information management and financial data analytics.
He holds a Ph. D from Indian Statistical institute, Kolkata and a certified associate of Indian Institute of Bankers.
He has published more than 30 papers in Indian and International scholarly journals. Presently, he is heading NMIMS Centre of Excellence in Analytics/Data Sciences as Director.