The present study attempted to analyse the random-walk characteristics of the gold spot price of the London Bullion Market Association (LBMA) by using several linear and nonlinear models. The research collects two decades of daily data from 3rd February 2000 to 2nd October 2020. Econometric tests such as serial correlation test, unit-root tests, multiple variance ratio (MVR), and the BDS test were applied to examine the linear and nonlinear dependence of return series. Further, we employed all the tests from ARCH family to examine the volatility clustering of the gold return series. The results of serial correlation and the unit-root test suggest that the gold return is stationary, therefore does not follow the random-walk benchmark, and hence the gold market is inefficient. EGARCH results indicate that the positive news has a more significant impact on the gold return than the negative news. The findings have important implications for the efficient portfolio investments, and better hedging opportunities for the investors.

本研究试图通过使用几种线性和非线性模型来分析伦敦金银市场协​​会（LBMA）黄金现货价格的随机游走特征。该研究收集了2000年2月3日至2020年10月2日二十年的每日数据。采用序列相关检验、单位根检验、多重方差比（MVR）和BDS检验等计量经济学检验来检验线性和非线性相关性回归系列。此外，我们采用了 ARCH 系列的所有测试来检验黄金回报序列的波动性聚类。序列相关和单位根检验的结果表明，黄金收益是平稳的，因此不遵循随机游走基准，因此黄金市场效率低下。 EGARCH结果表明，正面消息对黄金收益的影响比负面消息更为显着。研究结果对于有效的投资组合以及为投资者提供更好的对冲机会具有重要意义。