- Several strategies turned out to be profitable, although some of them in the very long term.
- ChatGPT also helped to perform backtesting automatically on TradingView.
In the field of bitcoin (BTC) and cryptocurrencies, trading has established itself as an activity that attracts investors in search of profitability.
Given the variety of trading strategies available, the question of which is the most suitable for each investment style and objective. At CriptoNoticias we have decided to analyze and compare four trading strategies proposed by the ChatGPT 4 artificial intelligence model.
In this report, we will present the results obtained by applying a scalping strategy, a day trading strategy, a futures-based daily candlestick strategy and, finally, a spot (spot) daily candlestick strategy. In all of them bitcoin will be traded against the most popular stablecoin, which is Tether USD (USDT) and indicators available in TradingView will be used.
Our purpose is to provide factual information that enables readers to make informed decisions in their bitcoin trading operations. Throughout the article, We will examine the advantages and disadvantages of each approach and their potential profitability. in different market contexts.
To test the effectiveness of the strategy, we will carry out what is called “backtesting”. This is testing how it would have turned out if applied to past operations.
Bitcoin scalping strategy
scalping is a very short-term trading strategy that is characterized by making multiple quick and small trades throughout a market session. The main objective of this technique is to make a profit by exploiting the minimum fluctuations in the prices of assets, in this case, bitcoin.
Scalpers, as traders who apply this strategy are known, seek to take advantage of price variations in very short times, such as seconds or minutes, in order to accumulate small profits that, at the end of the day, can represent a significant sum. To do this, they usually use technical analysis tools and price charts in very short intervals, such as 1, 5 or 15 minutes.
On this occasion, we asked ChatGPT 4 to develop a profitable strategy for scalping the BTC-USDT pair and this was their response:
In order to evaluate the effectiveness of this strategy, we ask ChatGPT 4 to write a script in Pine Script (TradingView programming language), to be able to do the backtesting on the trading platform itself, through its strategy simulator. The process was straightforward and required a couple of code fixes (which the AI app itself did).
Once the information was loaded in the strategy simulator, we proceeded to see the results. It is worth clarifying that the backtesting was done with the free version of TradingView, which allows the strategy to be applied to a maximum of 10,000 candles (in this case, 5-minute candles). As a consequence, the strategy was applied for 17 days, and in that period a total of 9 operations were executed. These were the results:
- Initial capital invested: 10,000 USDT.
- Number of operations closed: 9.
- Return of investment: -303.74 USDT (-3.04%).
In this way it is evident that the strategy, at least for the period of time evaluated, was not profitable.
Bitcoin day trading strategy
For those who want to take trading in a more relaxed way than scalpers, we are now looking for a day trading strategy.
day trading ands an investment strategy in which a trader buys and sells financial assets (in this case, bitcoin) within the same day of negotiation. The main objective of day trading is to make a profit by taking advantage of price fluctuations that occur during the hours that you are actively trading.
Unlike scalping, which focuses on very short-term trades (seconds or minutes), day trading usually covers operations that can last from a few minutes to several hours. At the end of the trading day, day traders generally close all their positions, avoiding holding them open overnight to reduce the risks associated with unforeseen events or changes in market conditions.
This is the strategy that ChatGPT 4 gives us:
The next step is to verify the effectiveness of the strategy, for which —as in scalping— we use the TradingView strategy simulator from a script generated by Chat GPT 4.
As we use 15-minute candles, the time period in which the strategy is evaluated is extended, reaching almost 2 months, with the following results:
- Initial capital invested: 10,000 USDT
- Number of operations closed: 23.
- Return of investment: 2,015.04 USDT (20.15%).
Despite the fact that in the first days there was an unrealized loss of considerable magnitude, at the end of this period, the strategy proved to be profitable.
Trading strategy with daily candlesticks in bitcoin futures market
Salping and day-trading have in common that they require constant attention from the trader. This must be in front of the computer screen viewing the bitcoin price chart to decide when to open or close operations.
For those who do not want to trade like this, we ask GPT to generate a futures trading strategy (just like the two presented so far), but with daily candles. In that way, the trader will be able to operate in longer periods of timeexecuting orders only at the close of the candles (which occurs at 0:00 Greenwich Mean Time).
Something to mention here is that ChatGPT 4 gave us, initially, three strategies that did not produce any entry orders in backtesting. It was only after the fourth attempt (and after solving several code errors, which involved several hours of dialogue with ChatGPT) that we came up with a strategy for this temporality, which is the following:
By using daily candles, we can backtest several years. Here we use it with the chart of the BTC-USDT pair on Binance, which takes us all the way back to 2017, when the exchange was founded. In that period, These are the results:
- Initial capital invested: $10,000
- Number of operations closed:78.
- Return of investment: 573.19 USDT (5.73%).
The strategy proved profitable. But it took 6 years to get a very small benefit, so It doesn’t seem to be convenient to implement.
Bitcoin spot trading strategy with daily candles
Spot trading, also known as spot trading, is a transaction in which the financial asset is bought or sold directly (in this case bitcoin). Unlike previous strategies, you will not trade a derivative here.
For practical purposes, this means that you can only earn with the rise of BTC and not with its fall, that is, you cannot invest in short (or “short”). The strategy to make money here, then, is to buy BTC at one price, and sell it at a higher price.
One advantage of this type of trading is that we will not pay maintenance fees for having open orders. Nor do we run the risk of being liquidated, that is, of losing the invested capital.
This is what ChatGPT 4 tells us to do:
When trying to do backtesting in the TradingView strategy simulator, the problems began. “The entered strategy does not produce any entry in the evaluated range,” the platform indicated.
For this reason, ChatGPT 4 had to be asked to make the necessary adjustments to solve this problem. The artificial intelligence application made adjustments on four occasions and, each time, it simplified the strategy, each time with fewer indicators. Finally, it turned out a strategy based solely on the moving average crossover (perhaps the most basic and well-known method of trading), as seen below:
And how does this strategy fare if we put it to the test? Let’s sees the backtesting results:
- Initial capital invested: $10,000
- Number of operations closed: 5.
- Return of investment: 15,291.57 (152%)
The strategy is profitable. But, as in the spot futures strategy with daily candles, it took 6 years to get this 152% profit. If the person, instead of trading bitcoin, had simply held his coins during that time, today he would have profits of around 1,000% due to the increase in the price of BTC.
Conclusions and final considerations
At CriptoNoticias we have explored and analyzed various bitcoin trading strategies proposed by ChatGPT 4, an artificial intelligence model developed by OpenAI. Throughout the analysis, we have found that ChatGPT can be a useful tool for traders in the search for profitable strategies and the design of their own investment approaches.
It is important to note that in these analyzes the commissions of the exchanges have not been taken into account, which could affect the results obtained in each case. Therefore, it is essential to take this aspect into account when implementing the strategies in practice.
It should also be considered that, if it is operated manually, it will be practically impossible for a trader to open and close all the operations indicated by the strategy. The backtesting is carried out considering that the trader operates 24 hours a day and 7 days a week.but, in reality, this fact does not occur.
Another thing to mention is that although we have tested only one strategy for each trading style here, ChatGPT could give us infinite different strategies, combining various indicators. Perhaps many of them are more effective than those discussed here.
Although the strategies presented in this article may be valid, the trader must complement them with technical analysis, fundamental analysis and other tools for better results. There is no single and perfect strategy, so it is necessary to adapt them to market circumstances and the objectives of each investor.
Finally, it is crucial to remember that nothing written here is investment advice. Cryptocurrency trading carries risks and each investor must take responsibility for their decisions and operations. It is always advisable to research thoroughly and have a well-structured investment plan to minimize risks and maximize the chances of success.
#tested #bitcoin #trading #strategies #ChatGPT #results