Background and Motivation

The subject of whether electricity markets are informationally efficient has received a lot of scholarly interest since the big European liberalisations of the 1990s. According to the Efficient Market Hypothesis (Fama, 1970), prices should represent all available information, making systematic profit from previous price data unfeasible. Samuelson (1965) provided the mathematical foundation with the martingale property, whereas Behavioural Finance (Thaler, 1999; Shiller, 2003; Kahneman & Tversky, 1979) challenged the rationality assumptions underlying EMH, arguing that efficiency can deteriorate under extreme market conditions.

However, electricity markets vary fundamentally from financial markets in that non-storability inhibits cross-temporal arbitrage, high production concentration allows for strategic price manipulation, and significant seasonal and intra-day patterns produce predictable price cycles. These characteristics produce kinds of informational inefficiency that are long-lasting and structurally anchored, rather than reflecting transitory behavioural biases.

Methodology

This study does a systematic literature review (SLR) in accordance with PRISMA criteria. A thorough search of Scopus, Web of Science, and Google Scholar for topics relating to electricity market efficiency, EMH, Variance Ratio Tests, Hurst exponents, GARCH models, and the Greek electricity market produced 312 suitable publications. After screening titles and abstracts (n = 87) and assessing full-text against pre-specified inclusion criteria (peer-reviewed empirical studies, 2002-2024, with at least one efficiency test applied to wholesale electricity or closely similar energy markets), 40 papers were selected for synthesis.

Key Findings

International evidence. Across geographies and approaches, the research agrees on one key finding: energy markets demonstrate systemic informational inefficiency. Weron (2002) used Hurst analysis to show a substantial mean reversion in NordPool pricing. Higgs and Worthington (2003) confirmed inefficiency across Australian markets using VRT, Morales and Hanly (2018) found gradual efficiency improvements in three European markets following regulatory integration under EU Directive 2009/72/EC, and Yang et al. (2009) are the rare exception, finding weak-form efficiency in Nordic electricity futures. Lin, Hsu, and Huang (2014) express a significant methodological worry about the possibility of data snooping when performing numerous tests to similar datasets, which is addressed by multi-indicator techniques such as the Electricity Market Efficiency Index (EMEI) created by Kristoufek and Vosvrda (2013).

The Greek market. The Greek power market is the most extreme known example of information inefficiency in Europe. Papaioannou et al. (2018) used SARMAX/GARCH to show that there were no leverage effects in Greek wholesale pricing from 2005 to 2016, indicating profound market failure. Papaioannou et al. (2019) used the EMEI to rank four European markets, and the Greek market came in bottom (VR = 0.319, H = 0.183, HHI = 7,739), compared to NordPool levels of VR = 0.507 and H = 0.361. The Public Power Corporation's dominating position (~85% market share) and lack of a competitive spot market design contributed to high inefficiencies.

Synthesis and divergence. The review finds a strong consensus: efficiency is dynamic, institutionally influenced, and inversely connected with market concentration. Divergences include the net effect of renewable energy penetration on efficiency, which reduces HHI (a positive effect) while introducing predictable intra-day price patterns such as the solar duck curve (a negative effect), as well as the sensitivity of results to the choice between daily and hourly data frequencies.

Research Gap

There is a critical gap in the literature: no systematic study examines the evolution of Greek electricity market efficiency following the November 2020 launch of the Hellenic Energy Exchange (HEnEx) under the Target Model, while simultaneously accounting for the unprecedented energy crisis of 2021-2023 (MCP peaking at €942/MWh in August 2022) and the rapid increase in renewable energy penetration (from ~32% to ~43% of demand between 2020 and 2023). Importantly, no research applies a battery of indicators — Variance Ratio Tests, Hurst Exponents (R/S, DFA, GPH, GHE), Sample Entropy, Fractal Dimension, GARCH/EGARCH, and HHI — to this era and market.

Ongoing Research

This evaluation informs an ongoing PhD project at Hellenic Open University, which uses an improved EMEI framework to analyse 49,386 hourly MCP data from HEnEx DAM (November 2020 - June 2026). Five research hypotheses are tested: (H1) whether market efficiency improved after the Target Model; (H2) whether HHI is negatively associated with efficiency; (H3) whether RES penetration reduces concentration but increases calendar-driven price patterns; (H4) whether the 2021-2023 energy crisis caused a temporary efficiency deterioration; and (H5) whether the Greek market is converging to NordPool efficiency benchmarks after 2020. A first descriptive examination of the dataset supports the structural breakdowns associated with each sub-period, laying a solid empirical foundation for further hypothesis testing.

JEL Classification: G14, Q41, Q47, L94, C22