As Bitcoin, and more globally cryptocurrencies, are getting more and more traction, we took a deeper look to try to predict bitcoin’s price. We gathered with our quant team and started to build a model.
First, we based our analysis on the similarities we could expect between bitcoin and other kind of assets. We compared bitcoin with :
a currency (main current usage), like the USDa commodity (to tokenize assets or execute smart contract), like oil for instancea safe-haven investment, like gold or silvera share, representing an investment in a community
Second, we tried to find a correlation between traditional drivers for pricing and bitcoin's price. As any other asset, the value is depending on supply and demand. But, for bitcoin, we took into consideration also:
number of merchants accepting bitcoininterest of participants, especially miners (linked to the game theory equilibrium)economic events (Brexit, etc.)security failures
Last, we used a time series model to try to predict the price, with first an ARIMA model. We also looked at the direct correlation between the number of addresses and the price (supposing the value of bitcoin is linked to the size of the community, following the Metcalf law). We finally also tried a GARCH model to handle the high and non-constant volatility.
Download full white paper here : http://www.slideshare.net/PatrickBucquet/bitcoin-pricing-jan2017