This paper analyzes the Ethereum blockchain and the applicability of models for network valuation in explaining the value of Ether. After examining the fundamentals of the Ethereum blockchain and its associated cryptocurrency, Ether, it is argued that the value of Ether is driven by the demand for computing power, but also that Ethereum shares many similarities with non-blockchain networks. Thus, the analysis applies the network valuation models by Metcalfe (2013), Briscoe, Odlyzko, and Tilly (2006), and Alabi (2017). While these models are based on the number of users, the findings suggest that other measures better represent demand of computing power on Ethereum. Thus, besides daily active IP addresses on Ethereum, the daily number of transactions and daily Gas used are applied as predictors. Contrary to previous findings, the analysis shows that Metcalfe’s Law and Alabi’s model are not able to explain the value of Ether over time. Metcalfe’s Law best explains the early value of Ether, and this paper suggests a combined model of Metcalfe’s Law and the model proposed by Briscoe, Odlyzko, and Tilly. The combined model proves to better explain the value of Ether over time compared to the single equation models. No evidence supports that Gas and transactions provide better predictors compared to active IP addresses. As the value of Ether seems to follow the proposed model, the model is used for forecasting to examine the predictability of Ether. The results show that existing models for forecasting activity and the demand for computing power on Ethereum challenge the predictability of the value of Ether. Thus, the final 6-month ahead forecast provides a wide range, and the value of Ether is forecasted between $96.95 billion and $132.34 billion for July 1, 2018. The forecast implies a gain between 35.36% and 84.78% over six months from December 31, 2017.
