# How is the banknote denomination designed?

Recently India demonetized their 500 and 1000 rupee notes. As a replacement a new 2000 rupee note was introduced. This has resulted in some chaos in the commercial markets due to decreased availability of banknotes. I was wondering about the root-causes of this.

I'm not a trained economist. One reason, that I think, adding to the chaos is the unscientific design of the denominations. Currently the banknotes in India have the following denominations: 5, 10, 20, 50, 100, 2000.

In the current price structure in India, there are lot of goods currently sold in the range 100-2000 and this means that for a given sale of a good in this range the money exchange process becomes very complicated due to the absence of some intermediate denominations in this large range.

For e.g: If somebody buys an item of 800 rupees, now she has to carry at least 8 different notes. On the other hand, if she decides to carry only a single note of 2000 rupee, then the shop owner has to give her 12 different notes back. So, either way now there is a large burden on the central bank to increase the number of notes in the supply.

So, I was wondering on what is the most efficient design of banknote denominations. efficient => The most optimizing in terms of the number of notes to carry/keep for the payer and the receiver and most optimizing for the central bank to issue in terms of cost and distribution logistics.

My intuition says that a denomination structure based on the powers of 2 would be the most 'efficient' mechanism. Another option would be the Fibonacci series. I would assume that most of the central banks across the world design their banknotes in a series close to the Fibonacci series. I tried to google around about how this is designed in general and surprisingly my searches didn't reveal anything in this direction.

My question is : Is my intuition right? Also it would be helpful if somebody points me to some literature discussing this topic.