Forecasting the Price of Gold. 2015.
This paper assesses the effectiveness of 17 forecasting models for gold prices. These models use monthly gold price data as well as Silver, Platinum, Palladium and Rhodium in some cases.
None of the models are able to forecast monthly gold prices over the long and the short run. Univariate models preformed better than the more complicated multivariate models developed in the paper. The most effective model was the Exponential Smoothing model. It reports the smallest forecast errors at 3, 4 5, 6, 7 and 8 months out. It outperforms the other models by betwee2n 22% and 84% based on the authors work.
It should be noted however that the best model to forecast the price next month in the Random Walk model. This means that over a short horizon gold prices are unpredictable.
The paper does not develop a trading rule to test the models. This would show if it is possible to make a profit trading gold based on any the forecasts in methods. Without this it is impossible to say whether this paper shows a way to beat the gold market.
It is also unclear why only monthly data is used as daily data is freely available for all of the metals used in the study.
Data: Monthly observations January 1972 – December 2013
This article seeks to evaluate the appropriateness of a variety of existing forecasting techniques (17 methods) at providing accurate and statistically significant forecasts for gold price. We report the results from the nine most competitive techniques. Special consideration is given to the ability of these techniques to provide forecasts which outperforms the random walk (RW) as we noticed that certain multivariate models (which included prices of silver, platinum, palladium and rhodium, besides gold) were also unable to outperform the RW in this case. Interestingly, the results show that none of the forecasting techniques are able to outperform the RW at horizons of 1 and 9 steps ahead, and on average, the exponential smoothing model is seen providing the best forecasts in terms of the lowest root mean squared error over the 24-month forecasting horizons. Moreover, we find that the univariate models used in this article are able to outperform the Bayesian autoregression and Bayesian vector autoregressive models, with exponential smoothing reporting statistically significant results in comparison with the former models, and classical autoregressive and the vector autoregressive models in most cases.