黃金相關類股價格預測黃金價格

Abstract

本研究以黃金相關類股價格加權指數作為主要的預測變數，以美國、加拿大、澳大利亞等國的月資料為樣本，檢視黃金相關類股價格對黃金現貨價格是否具有預測能力。透過樣本內與樣本外預測的檢定結果顯示，各國黃金相關類股價格對黃金現貨價格的預測能力，皆可擊敗不變預測模型。我們也進一步檢視黃金相關類股價格與其餘控制變數的預測能力，發現黃金相關類股價格對黃金價格的預測能力皆優於通貨膨脹率、利率、匯率以及油價等變數。相較於過去較常用於建立金價預測模型的變數，由於股價在市場資訊中具有領先的特性，我們也發現黃金相關類股價格可以增加模型的解釋能力，但無法顯著提升模型的預測能力，其可能主因為在黃金相關類股中，組成多為礦產企業，對金價的影響多聚焦在供給面，而通貨膨脹、利率、匯率以及油價等變數則於供給與需求面皆有影響，故這些控制變數對黃金價格的解釋能力可能已包含部分黃金相關類股的解釋能力。

This study employs the gold-related equity price-weighted index as the main explanatory variable and monthly data from countries such as the United States, Canada, and Australia as the sample, in order to examine whether the gold-related equity price has predictive power for the spot gold price. The testing results of in-sample and out-of-sample forecasting demonstrate that the predictive ability of the gold-related equity price for the spot gold price can outperform the no-change forecast model. We further examine the predictive abilities of gold-related equity prices and other control variables and find that the prices of gold-related equity outperform variables such as inflation rates, interest rates, exchange rates, and oil prices in predicting gold prices. Compared to variables commonly used in constructing gold price forecasting models, such as inflation rates, interest rates, exchange rates, and oil prices, stock prices possess leading characteristics in market information. It is found that incorporating gold-related equity prices into the model can enhance the explanatory power of the model, but can not significantly improve its predictive ability. This may be mainly because gold-related equity mainly consist of mining companies, and their impact on gold prices is focused on the supply side, while variables such as inflation rates, interest rates, exchange rates, and oil prices have an impact on both the supply and demand sides. Therefore, the explanatory power of these control variables for gold prices may already include some of the explanatory power of gold-related equity. Furthermore, the impact of the inclusion of gold-related equity prices on the explanatory power of the gold price forecasting model is examined, and it is found that it is not a key explanatory variable. This is possibly because the gold-related equities are mostly composed of mining enterprises, and their impact on gold prices is mostly focused on the supply side, while variables such as inflation and interest rates have a greater impact on the demand side. Therefore, although gold-related equity prices are not a key explanatory variable, they can increase the explanatory power of the model.