### Visualizing the relationship between Cryptoasset Price and Supply

One of the main advertised advantages of cryptocurrencies is the limited supply. Supply and demand is a well known economic model that helps in determining the price of an asset. It is a popular mantra that the more limited the supply is, the more probable demand can get, and thus increase in price. We can see the same behavior in cryptocurrencies. The interactive plot below demonstrates this behavior:

The chart plots price in USD (y-axis) vs available supply (x-axis) for all of the 1633 cryptocurrencies from Coinmarketcap as of 29 May 2018. Both of the x-axis and the y-axis are represented in log-scale. The size of each data point reflects the coin’s rank as per market cap.

From the interactive plot, we can observe several points:

1. The majority of coins exist within the ranges of supply between 1 million and 1 billion coins. Also, most of them have a price tag range between 0.01–1 USD.
2. As expected, the price tends to decrease as supply increases. However, what is more interesting is to realize the rate of the price drop with respect to supply increase. Can we find a function that articulates this relationship?

In data analysis, we have a common analysis tool called regression which tries to estimate the correlation between variables — in this case, price and supply — by fitting-in the input data into an output function. We ran several regression tools on the underlying data and found that the power function below fits the most, where P is price and S is Supply.

The function resembles the purple line within the plot. Note that this is NOT a straight line as it is plotted in a log-log chart.
The interesting point here is not the constant 2669.7, but the relationship between P and S. First, they are inversely proportional to each other, which makes perfect sense. Secondly, which is the more interesting part, is that the function is not exactly inversely proportional to supply.

#### In other words, as the supply of a coin increases, the decay in its price decreases.

This means that the Price (P) line decays fast at the start but then the line plateaus fast with very minimal change, as supply (S) increases. This suggests that increasing supply does not have a strong effect on Price after a certain threshold.

We tried to validate this observation on different clusters of data. Rather than fitting a function on the entire dataset of 1633 coins, we tried to fit it into different coin categories, and the results show a similar trend. Here are two charts that demonstrate this behavior.