The Nature of Trends and Cycles of Ethereum

Abstract

The chapter examines the nature of long-run nonlinear trends of the closing price of Ethereum in terms of USD, from 2015m 08 to 2023m 05 using the econometric model of the Box and Jenkins’ ( 1976 ) methodology of ARIMA (p, d, q) and Hamilton (2018) decomposition. Additionally, the forecast behaviour for 2025m 01 was computed with/without the Hamilton regression filter. The automatically selected ARIMA model of Ethereum price is convergent, and its forecast path for 2025m 01 showed insignificance with seasonal fluctuations. However, its decomposition model is cyclical, cyclically trending, and seasonally fluctuated, with forecast behaviour that is convergent, stable and significant without seasonal variation.