Equilibrium. Quarterly Journal of Economics and Economic Policy Volume 15 Issue 2 June 2020 p-ISSN 1689-765X, e-ISSN 2353-3293 www.economic-policy.pl ORIGINAL ARTICLE Citation: Hanias, M., Tsakonas, S., Magafas, L., & Thalassinos, E. I., & Zachilas, L. (2020). Deterministic chaos and forecasting in Amazon’s share prices. Equilibrium. Quarterly Journal of Economics and Economic Policy, 15(2), 253–273. doi: 10.24136/eq.2020.012 Contact to corresponding author: stef1920@hotmail.gr; Department of Economics, University of Thessaly, 28hs Octovriou 78, Volos, P.C. 38333, Greece Received: 28.02.2020; Revised: 4.04.2020; Accepted: 27.04.2020; Published online: 24.06.2020 Michael Hanias International Hellenic University, Greece orcid.org/0000-0002-0918-9343 Stefanos Tsakonas University of Thessaly, Greece orcid.org/0000-0003-2878-7741 Lykourgos Magafas International Hellenic University, Greece orcid.org/0000-0002-8253-9653 Eleftherios I. Thalassinos University of Piraeus, Greece; University of Malta, Malta orcid.org/0000-0003-3526-4930 Loukas Zachilas University of Thessaly, Greece orcid.org/0000-0001-7021-2955 Deterministic chaos and forecasting in Amazon’s share prices JEL Classification: C53; C63; G17 Keywords: time series; chaos theory; econophysics; forecasting Abstract Research background: The application of non-linear analysis and chaos theory modelling on financial time series in the discipline of Econophysics. Purpose of the article: The main aim of the article is to identify the deterministic chaotic behav- ior of stock prices with reference to Amazon using daily data from Nasdaq-100. Methods: The paper uses nonlinear methods, in particular chaos theory modelling, in a case study exploring and forecasting the daily Amazon stock price. Findings & Value added: The results suggest that the Amazon stock price time series is a deter- ministic chaotic series with a lot of noise. We calculated the invariant parameters such as the maxi-mum Lyapunov exponent as well as the correlation dimension, managed a two-days-ahead forecast through phase space reconstruction and a grouped data handling method.