The Trading Algorithm:

Our analysis employs a straightforward trading strategy that issues buy and sell signals based on two primary components: the STORSI indicator and two EMAs. The STORSI indicator gauges the momentum of a financial asset by comparing its closing price to the high-low range over a designated timeframe. The EMAs are computed based on the average price of the asset during a specified period, with a greater emphasis on recent prices.

Here’s a simplified version of the trading algorithm code:

signals.append(0)

for i in range(1, len(df)-1):

# Define trading strategy

if df.iloc[i]['STORSI'] > 0.2 and df.iloc[i]['EMA5'] > df.iloc[i]['EMA6']:

signals.append(1)

elif df.iloc[i]['STORSI'] < 0.8 and df.iloc[i]['EMA6'] > df.iloc[i]['EMA5']:

signals.append(-1)

else:

signals.append(0)

signals.append(0)

The code initializes an empty list for signals and iterates through the rows of a pandas DataFrame df. For each entry, it evaluates two conditions: if the STORSI exceeds 0.2 and the 5-period EMA is greater than the 6-period EMA, a buy signal is created (adding a value of 1 to the signals list); if the STORSI is below 0.8 and the 6-period EMA surpasses the 5-period EMA, a sell signal is produced (adding -1), otherwise, it appends a 0 for no signal.

The algorithm also incorporates two extra zeros at both the beginning and the end of the signals list to maintain a neutral starting and ending position.

Backtesting the Strategy:

To assess the algorithm's effectiveness, we backtested it using historical price data of a specific financial asset. The backtesting process is depicted in the following code snippet:

df["signal"] = signals

print(signals)

investment = 1000

current_investment = 1000

# Remaining code omitted for brevity...

We tested the algorithm using ETHUSDT price data over an hourly timeframe, analyzing profits generated by various EMA period combinations. After experimenting with combinations ranging from 20 to 100, we discovered that the most effective combination was EMA27 and EMA45, yielding a profit of $280,491. In contrast, the EMA28 and EMA48 combination resulted in a profit of only $172,426.

The following chart illustrates the performance variations of the algorithm based on different EMA period selections:

Discussion of Results:

The backtesting findings indicate that selecting EMA periods significantly impacts the profitability of the trading strategy. The optimal EMA combination of EMA27 and EMA45 generated a profit of $280,491, while the less effective EMA28 and EMA48 combination produced only $172,426—a notable difference of nearly $108,000 or 63%.

This disparity arises because the optimal EMA combination more effectively captures the asset's trend. The 27-period EMA is more responsive to recent price shifts than the 28-period EMA, whereas the 45-period EMA better reflects long-term price trends compared to the 48-period EMA. This sensitivity allows the optimal combination to swiftly react to short-term price movements while also accounting for the overall asset trend.

It's essential to recognize that the choice of EMA periods can vary based on the specific asset in question and the timeframe being analyzed. Therefore, conducting thorough testing and analysis is crucial to identify the best EMA periods for a given asset and timeframe. This might involve experimenting with different EMA combinations alongside adjusting additional trading algorithm parameters such as STORSI thresholds, stop-loss, and take-profit levels.

In conclusion, the selection of EMA periods plays a pivotal role in the profitability of an algorithmic trading strategy. Our backtesting results suggest that the optimal combination of EMA27 and EMA45 can yield substantially higher profits compared to less effective combinations like EMA28 and EMA48. However, the ideal EMA periods may differ based on the asset and timeframe under consideration, underscoring the need for comprehensive testing and analysis to refine a trading algorithm's parameters.

By diligently fine-tuning EMA periods, traders can enhance the effectiveness and profitability of their algorithmic trading strategies.