The Textbook for AI Stock Investing Aiming for 1 Million Yen per Month
“If you use AI, you can earn 1 million yen a month from stocks.”
When you see a claim like that, you might get your hopes up a little.
Have ChatGPT analyze stocks and buy the ones that look like they will go up. AI can process massive amounts of information in a short time and seems capable of making more rational decisions than a human.
If that’s the case, isn’t there no need to stare at charts for hours on your own?
I think there are people who think that way.
However, there is a major pitfall here.
Just because AI is excellent does not mean you will make a profit in stock investing.
In fact, I believe that people who trade by trusting AI’s answers exactly as they are are the ones in the most danger.
Even so, there is plenty of value in using AI for stock investing.
This is because the true strength of AI lies not in “predicting future stock prices,” but in organizing the information necessary for investment decisions, creating trading rules, and reducing judgment errors.
The reality you should know before aiming for 1 million yen per month
First, I would like to think about the goal of 1 million yen per month in terms of numbers.
Suppose you want to earn a stable profit of 1 million yen every month; how much investment capital would you need?
| Monthly Profit Rate | Capital Required for 1 Million Yen/Month |
| —– | ———— |
| 1% | 100 Million Yen |
| 2% | 50 Million Yen |
| 5% | 20 Million Yen |
| 10% | 10 Million Yen |
| 20% | 5 Million Yen |
*Simple calculation without considering taxes and fees. Actual profit rates are not constant, and there will be months where you incur losses.
You can see it when you look at this.
To earn 1 million yen every month with 1 million yen in capital, you would need a monthly return of 100%. It is hard to call that a realistic investment plan.
On the other hand, if you think in terms of increasing your capital while improving not just your win rate but also your profit-loss balance, you can create a concrete plan.
However, even achieving a monthly return of 5% or 10% consistently is not easy.
That is why I think it is more rational to build a system that makes it hard to lose capital and then think about how to accumulate profits, rather than looking for a way to earn 1 million yen a month from the start.
The first thing you should change in AI stock investing
For example, suppose you ask ChatGPT the following:
“Tell me which Japanese stock will rise the most tomorrow.”
There is a problem with this way of asking.
No matter how plausible the stocks the AI suggests are, you won’t know the basis for the judgment or how to respond if the prediction is wrong.
So, let’s try changing the question a little.
Even just doing this will significantly change how you use AI.
Instead of just having stocks suggested to you, you will be able to think about why you are buying them and at what point you should judge that you were wrong.
Of course, an environment where the AI can access the latest data is necessary, and checking the outputted prices and financial results is also essential.
Even so, it is effective as a foundation for building investment decisions.
There is a number more important than win rate
When trying to make a profit in stocks, you tend to worry about “how many times you won.”
But profit is not determined by win rate alone.
For example, let’s consider the case of trading 10 times.
-
7 wins, with a profit of 10,000 yen each time
-
3 losses, with a loss of 30,000 yen each time
Even though the win rate is 70%, the total is a loss of 20,000 yen.
Conversely, even if the win rate is 40%, if the profits sufficiently exceed the losses, there is a possibility of being in the black.
In other words, what you should seek from AI is not a simple prediction accuracy rate.
Find trading conditions with a positive expected value and verify whether they hold true in actual trading.
I believe this is the starting point for AI stock investing.
The mechanism needed to aim for 1 million yen per month
If you are seriously considering AI stock investing, you need at least the following flow.
-
Narrow down candidate stocks for trading using AI
-
Verify trading conditions based on past stock prices and trading volume
-
Determine criteria for entry, stop-loss, and profit-taking
-
Record actual trading results with a small amount of capital
-
Analyze results with AI and improve the rules
What I especially do not want you to overlook are points 4 and 5.
How did the stocks that the AI judged as having a ‘high probability of rising’ actually perform?
Were you able to capture the expected profit? Did the stop-loss function as intended? Did a profit remain even after considering commissions and slippage?
Unless you record up to this point, you will not know if that investment method is truly effective.
And I think the goal of 1 million yen per month should be placed after building such a mechanism. It is not a guaranteed figure, and its feasibility changes depending on the necessary capital and the losses you can tolerate.
From here on, I will break down into a practical format everything from specific prompts for stock selection using ChatGPT, how to create trading rules, methods for verifying trading results, and designing profit targets according to your operating capital.