In this article, we will analyze a peculiarity of the cryptocurrency market, namely the correlation that exists between the price trend of Bitcoin (BTC) and that of Ethereum (ETH), the main altcoin in the market. We will try to leverage this statistic to create a systematic trading strategy, working on the Ethereum-Bitcoin (ETH-BTC) pair.
The characteristic that one wants to analyze is linked to an alleged recurring behavior (otherwise known as Bias) that would manifest itself during the course of the week: one wants to verify if it is true that it is convenient to hold Bitcoin (BTC) from the beginning of the week until the weekend, and convert it into Altcoin (therefore into Ethereum in this example) during the weekend, when indeed Bitcoin seems to take a break in favor of the Altcoin which instead appreciate.
The concept of bias, meaning a recurring behavior of prices over time, is one of the simplest triggers on which it is possible to build a trading strategy. It is a market inefficiency that reoccurs with a certain regularity, for example, in specific time slots or days of the week. When it extends to longer periods, it is usually referred to as seasonality.
Correlation between Bitcoin and Ethereum: price analysis over time
To verify if this recurring behavior between Bitcoin (BTC) and Ethereum (ETH) truly exists, the Bias Finder will be used in the preliminary phase, a tool developed internally at the Unger Academy to simplify the search for biases on futures. The Bias Finder, in fact, based on the historical data of a certain instrument, can quickly analyze trends on candles with a variable duration between 5 and 60 minutes. Once the instrument and the time frame are selected, it is then possible to evaluate different time horizons: daily (Intraday), weekly (Weekly), monthly (Monthly), or annual (Yearly). For each of these, a chart is generated that shows the trend of the selected instrument, expressed as the average monetary excursion over time in absolute terms or as a percentage.
In our case, by selecting the Ethereum-Bitcoin (ETH-BTC) pair, with a 15-minute time frame and a data history from 01/01/2018 to 31/12/2024 (data source: Binance), one can quickly assess whether the behavior that was assumed during the week is present or not, by analyzing the trend of the average monetary excursion of the instrument.
Figure 1 – Graphical representation of the weekly bias between Ethereum and Bitcoin (ETH-BTC) from 2018 to 2024.
From the weekly chart shown in Figure 1, where each line represents a year (from 2018 to 2024) and the dark blue line represents the average for the entire period, we can see that statistically, excluding 2021 (plotted in green), the Ethereum-Bitcoin pair shows a certain tendency to decline (the price of Bitcoin increases compared to that of Ethereum) from the beginning of the week until about Thursday, then slightly rises during the weekend.
The bias, therefore, although not very pronounced, still seems to be present, so one could hypothesize a systematic trading strategy to evaluate its profitability.
Logic of the Bias trading strategy on the Ethereum-Bitcoin (ETH-BTC) pair
Initially, using a 60min chart of Ethereum-Bitcoin, and referring to the UTC time zone, one might hypothesize selling (i.e., buying Bitcoin) on Monday at midnight (night between Sunday and Monday), and buying Ethereum (ETH) on Friday also at midnight. Any open position, therefore, would be closed by opening the position in the opposite direction, without any stop loss or take profit order.
For simplicity in reading the results, operations will be conducted with 100 Bitcoin (BTC) for each trade, thus allowing a direct evaluation of the results as a percentage of the capital employed, which is precisely 100 Bitcoin (BTC). Figures 2 and 3 show the metrics of this very simple strategy, without any type of filter applied to the operations.
Figure 2 – Equity line of the base strategy on Ethereum-Bitcoin (ETH-BTC).
Figure 3 – Trade Analysis of the base strategy on Ethereum-Bitcoin (ETH-BTC).
It is immediate to notice how the growing equity confirms the presence of the bias identified with the Bias Finder, although the trend is rather volatile in the historical range considered.
Evaluating the average trade, a value of 0.38 Bitcoin (BTC) is found, which is 0.38% of the capital employed (100 Bitcoin), which could also constitute a value adequate to cover operational costs, considering that it is a rough strategy, but it can certainly be improved.
Optimization of the strategy on Ethereum-Bitcoin: techniques to improve performance
One could first optimize the market entry timing by varying the days and entry time: for the long entry day between Wednesday and Saturday, and for the short entry day between Sunday and Tuesday. Figure 4 shows the result of the optimization performed with MultiCharts, where the days of the week range from 0 (Sunday) to 6 (Saturday). It is noted that the long entry (purchase of Ethereum) on Wednesday at 16:00 is more effective compared to Friday, as previously hypothesized. The short entry can instead be left on Monday at midnight, since advancing or delaying it by an hour does not substantially change the result.
Figure 4 – Optimization of the days and times of market entry for the strategy on Ethereum-Bitcoin (ETH-BTC).
The next step could be to evaluate the use of stop loss and take profit to better manage open positions. By optimizing the stop loss between 0 and 5 Bitcoin (equivalent to 0% – 5% of the open position), in steps of 0.1, and the take profit between 0 and 10 Bitcoin in steps of 0.25, you get the combinations in Figure 5, among which you could opt for the one with Stop 1.9 Bitcoin and take profit 7.5 Bitcoin, which leads to an excellent Net Profit to Max Drawdown ratio and an average trade that rises to 0.55%.
Figure 5 – Optimization of stop loss and take profit values of the strategy.
As illustrated in Figure 6, the equity line has definitely improved, as have all the metrics of the strategy observed previously.
Figure 6 – Equity line of the strategy on Ethereum-Bitcoin (ETH-BTC) optimized.
However, by analyzing long trades and short trades separately, it can be seen that the long side is the less effective one, so one might consider filtering long entries using price patterns, to operate only when conditions are more favorable for the development of the bias that one is trying to ride. Nothing would prevent doing the same on the short side, provided that one does not filter too much and risk falling into overfitting the strategy.
Figure 7 – LONG and SHORT equity line of the strategy on Ethereum-Bitcoin (ETH-BTC) optimized.
Conclusions and insights for developing a trading strategy based on the Bitcoin-Ethereum correlation
The bias identified between Ethereum and Bitcoin represents an interesting example of market inefficiency that can be exploited with a systematic approach. Despite the simplicity of the initial approach, the results highlight significant potential, especially after optimizing the entry and exit rules of the trades.
This study aims to be a starting point for further exploring the topic and refining the strategy to further improve its performance. The opportunities for refinement are numerous, and the curiosity and initiative of the reader are left with the task of exploring them to create increasingly effective systems.
Until next time and happy trading!
Andrea Unger
Read the full article here
