Efficiency Dynamics of the Precious Metals Market: The AMIM Approach

Technical Summary: Efficiency Dynamics of the Precious Metals Market: The AMIM Approach

Problem Statement
This study addresses the need to understand the time-varying efficiency of precious metals markets (gold, silver, platinum, and palladium) amidst significant global economic disruptions. Specifically, it investigates how market efficiency shifted during the COVID-19 pandemic (December 9, 2019 – May 5, 2023) and the subsequent post-pandemic normalization period (May 8, 2023 – June 17, 2026). While the Efficient Market Hypothesis (EMH) posits that asset prices reflect all available information, existing literature suggests that market efficiency is not static but is influenced by regime shifts, financial turmoil, and geopolitical risks. The authors note that traditional static measures often fail to capture these dynamic inefficiencies, particularly in the context of safe-haven assets where investor behavior and information processing change rapidly during crises.

Methodology
The research employs the Adjusted Market Inefficiency Magnitude (AMIM) approach, a methodology developed by Tran and Leirvik (2019, 2020). This method was selected for its ability to provide robust, less biased estimates of time-varying efficiency compared to traditional measures, particularly when dealing with finite sample sizes.

• Data: The study utilizes daily closing prices for gold, silver, platinum, and palladium from "investing.com," spanning from December 9, 2019, to June 17, 2026.
• Return Transformation: The primary analysis uses simple returns ($r_{i,t}$) rather than logarithmic returns. The authors argue that logarithmic transformations can mechanically alter the distributional and autocorrelation properties of returns, potentially distorting the estimation of market inefficiency. Simple returns are used as the benchmark model, with logarithmic returns employed solely as a robustness check.
• Estimation Technique: The AMIM is calculated using overlapping rolling-window estimations. The process involves:
  1. Estimating an autoregressive model where coefficients ($\beta$) are tested for significance.
  2. Calculating the Market Inefficiency Magnitude (MIM) based on the sum of absolute standardized autocorrelation coefficients.
  3. Adjusting the MIM by subtracting a confidence interval range (RCI) and normalizing it against the theoretical maximum to derive the AMIM.
  4. An AMIM value of 0 or less indicates an efficient market, while positive values indicate inefficiency.
• Robustness: To ensure the stability of findings, the analysis was re-estimated using logarithmic returns. The authors note that while this transformation affects the statistical properties of the series (particularly for volatile assets like platinum and palladium), it serves to validate the core conclusions rather than compete with the baseline simple-return specification.

Key Contributions
The paper makes three primary contributions to the literature:

1. Methodological Application: It applies the AMIM methodology to precious metals in a fine-grained manner, offering a more robust framework for detecting time-varying inefficiencies than traditional autocorrelation or variance ratio tests.
2. Comparative Scope: Unlike studies focusing on a single metal or market condition, this research provides an expansive comparative analysis across four leading precious metals across two distinct regimes: the pandemic and the post-pandemic era.
3. Empirical Insight into Regime Dependence: It generates new empirical results regarding how efficiency adjusts differentially across assets in response to changing macroeconomic environments, deepening the understanding of adaptive market behavior under heightened uncertainty.

Empirical Results
The study finds that all four precious metals generally adhere to the weak-form Efficient Market Hypothesis throughout the entire sample period, as evidenced by AMIM values close to zero. However, significant heterogeneity exists across assets and time periods:

• Overall Efficiency Ranking: Over the full sample horizon, palladium is identified as the least efficient metal, followed by silver. Gold and platinum generally exhibit higher efficiency levels.
• Pandemic vs. Post-Pandemic Dynamics:
  • Gold: Demonstrated the highest efficiency during the COVID-19 pandemic but experienced a relative decline in efficiency in the post-pandemic period.
  • Silver: Is the only asset that exhibited an improvement in efficiency from the pandemic to the post-pandemic period. Consequently, silver became the most efficient asset in the post-pandemic era, overtaking gold.
  • Platinum and Palladium: Both experienced a relative decline in efficiency after the pandemic, though palladium remained the least efficient overall.
• Robustness Check: The logarithmic return analysis largely confirmed the main findings: all markets remained generally efficient, gold showed a decrease in relative efficiency post-pandemic, and silver consistently improved. However, the log-return specification suggested a slight improvement in relative efficiency for platinum and palladium post-pandemic, a divergence the authors attribute to the sensitivity of these volatile assets to return transformation methods.

Significance and Claims
The authors claim that their findings highlight the regime-dependent nature of market efficiency. The study asserts that market efficiency is not a constant property but varies across asset classes and over time as macro-financial conditions change.

• For Investors and Policymakers: The results suggest that investment, hedging, and diversification strategies must be adjusted to changing market conditions rather than relying on assumptions of constant efficiency. The identification of silver as an increasingly efficient asset in the post-pandemic era, and gold's relative decline, offers specific implications for portfolio allocation and risk management.
• Theoretical Implication: The findings reinforce the Adaptive Market Hypothesis, demonstrating that market efficiency is a function of economic uncertainty and structural change.
• Methodological Value: The study positions the AMIM framework as a promising alternative to static baselines, capable of capturing elusive, time-varying changes in market efficiency that traditional approaches might miss.

The paper concludes by acknowledging limitations, noting that the analysis is restricted to four metals and specific time periods, and suggests future research could extend to other commodity classes or combine AMIM with machine learning and multifractal techniques to better understand volatility spillovers and market connectedness.