Using the Holt-Winters Model for Trading Bitcoin and Gold

Investing in gold and bitcoin can be a great way to diversify your portfolio and protect your wealth. Gold is a physical asset that has been used as a currency for centuries. Bitcoin is a digital asset that was created in 2009. These assets have attracted the attention of traders and investors due to their volatility and potential for high returns.

Reference [1] proposed a method based on the Holt-Winters model [2,3] for trading gold and bitcoin. The Holt-Winters model is a statistical technique that is used to forecast time series data.  It utilizes exponential moving averages to smooth out the data and make forecasts.

The Holt-Winters model has three parameters:

– Alpha (α): controls the level of the time series,

Beta (β): controls the trend of the time series, and

– Gamma (γ): controls the seasonality of the time series.

The model is fit to data using the least squares method. The fitted model can then be used to make predictions about future values of the time series.

The authors pointed out,

The present study outlines a mathematical model and trading strategy that maximize the profits in gold-bitcoin trading. The only data used were the historical prices of gold and bitcoin. To avoid some limitations in previous study, both financial methods and statistical analysis were applied in this study. In financial aspect, indicators like relative strength index (RSI), momentum index and simple moving average (SMA) were computed. Trading shrink ratio was also calculated to control the trading volume. In statistical aspect, non-seasonal Holt-Winters’ Model was applied to predict price dynamics. A trading strategy was developed combing both financial and statistical information. The result indicates that traders can make huge profit by using our strategy. In addition, the sensitivity test suggests that the change of commission fees has little impact on our model.

Given the amount of research presented in the paper, we think that the authors have made a bold claim. We believe, however, that the idea of applying a predictive technique to trading has its merits, and it’s worth investigating how one can incorporate models such as the Holt-Winters into the trading framework.

Let us know what you think in the comments below or in the discussion forum.


[1] Yan Chenge, Wu Zebin, Gan Haoyu, Zhu Ziwen, Mathematical modeling and dynamic trading strategies for gold and bitcoin, Academic Journal of Mathematical Sciences, Vol. 3, Issue 1: 47-54

[2] Holt, C. E. (1957). Forecasting seasonals and trends by exponentially weighted averages (O.N.R. Memorandum No. 52). Carnegie Institute of Technology, Pittsburgh USA.

[3] Winters, P. R. (1960). Forecasting sales by exponentially weighted moving averages. Management Science, 6(3), 324–342.

Further questions

What's your question? Ask it in the discussion forum

Have an answer to the questions below? Post it here or in the forum


Leave a Reply