Challenging Japanese Stock Swing Trading with ChatGPT #55
I continue to engage in short-term swing trading, where I purchase Japanese stocks during my lunch break on weekdays and, in principle, sell them during the lunch break of the next business day. I use ChatGPT for stock selection and trading decisions, verifying with my own money whether it is actually possible to make a profit in Japanese stock trading using AI.
This article, excluding the final “Thoughts on this round” section, was also created by ChatGPT. The main text is organized based on actual trading records and the analysis performed at that time.
Profit on all 3 stocks for +34,990 yen! Yet, the cumulative loss remains at approximately 180,000 yen.
On Monday, November 16, I settled the three stocks I was holding and locked in a total profit of 34,990 yen. In addition to the margin short sale of Kioxia Holdings, the cash transactions for Bridgestone and Kajima Corporation also resulted in gains.
However, even with this profit, the cumulative profit/loss is -179,920 yen. Across 142 total trades, I have 76 wins and 66 losses, with a win rate of 53.5%. The difficulty of short-term trading is reflected in the numbers: even with more wins than losses, a loss remains.
Kioxia margin short sale successful. Kajima Corporation also yields a profit of 12,290 yen.
Kioxia Holdings (285A)
I bought back 100 shares on November 16 at 8,462 yen, which I had sold short on margin at 8,622 yen on November 9, locking in a profit of 16,000 yen.
At the time of the short sale, I focused on Kioxia’s relative weakness even while the market was rebounding. Although the direction anticipating a decline was correct in the end, I held the position longer than usual this time. It is necessary to distinguish this from trades where the prediction for the next business day was accurate.
Bridgestone (5108)
I sold 100 shares at 4,441 yen, which I had purchased on November 13 at 4,374 yen. The realized profit was 6,700 yen.
What I evaluated at the time of purchase was that the stock price had recovered from the morning decline following the earnings announcement. It remained above the purchase price even after the weekend, so the judgment on the upward direction was successful. However, it is impossible to judge from the execution price alone whether the strength of the upward trend was as expected.
Kajima Corporation (1812)
I sold 100 shares at 5,003 yen, which I had purchased on November 12 at 4,880.1 yen, locking in a profit of 12,290 yen.
In my purchase decision, I evaluated the strength of the stock price against the backdrop of an upward revision in earnings forecasts. The selling price exceeded the 5,000 yen milestone, and the decision based on the material led to a profit. On the other hand, since I held it beyond the next business day, it is necessary to verify separately until what point the effect of the material lasted.
*All profit/loss figures are before deductions for taxes, commissions, and margin-related expenses.
Three new stocks purchased. High hopes for Bridgestone again.
On November 16, I purchased 100 shares each of three stocks that ChatGPT had identified as purchase candidates with conditions.
Astellas Pharma (4503): 2,744.3 yen
In addition to the relative strength of the pharmaceutical sector, I evaluated the fact that the morning trading volume was approximately 1.38 times the comparison benchmark and the stock price was approximately 0.74% above the VWAP. This is my top candidate this time.
OBIC (4684): 6,264 yen
The reason for selection was that it did not involve a large gap-up, was raising its lows, and was trading about 0.65% above the VWAP. The goal is to avoid chasing overheated, rapidly rising stocks.
Bridgestone (5108): 4,449 yen
I repurchased a stock that I had taken profits on the same day. While I evaluated the continuation of the rise and the price movement being about 0.58% above the VWAP, concerns about demand due to downward revisions in earnings by competitors were also pointed out.
Note that for all three stocks, ChatGPT’s judgment was “conditional purchase.” The suggested maximum purchase prices were 2,758 yen for Astellas Pharma, 6,260 yen for OBIC, and 4,445 yen for Bridgestone, but the actual purchase prices for the latter two stocks were 4 yen higher in each case.
Next time: Verifying “prediction accuracy” and “adherence to purchase conditions”
Although I was able to take profits on all three stocks this time, because the holding periods differ, it cannot simply be treated as a success rate for next-day predictions.
Next time, I will check whether the three newly purchased stocks can maintain their rise until noon on the next business day. I am particularly focusing on the price movement of Bridgestone, which I repurchased after taking profits, and the performance of the two stocks that exceeded their maximum purchase prices.
Beyond just stock selection, to what extent should I adhere to the purchase conditions indicated by ChatGPT? I will continue to verify this based on actual execution results.
Thoughts on this time
・Arrivederci! Kioxia!
・I was finally able to part ways with Kioxia. A high-volatility stock like you is enough to handle in day trading.
・Since I was on paid leave today, I dabbled with Kioxia in day trading, but it was so sideways that there was almost no timing to touch it. It makes me nervous, my eyes get tired, and I could never become a full-time day trader.
*This article is a record of personal transactions and verification, and does not recommend the buying or selling of any specific stock.
#Investment #StockInvestment #JapaneseStocks #Stocks #AssetManagement #AssetFormation #SwingTrading #InvestmentRecord #IndividualInvestor #StockInvestmentBeginner #ChatGPT #AIInvestment #MarginTrading #InvestmentBeginner