Auto-Trading Week 1 — Position Rotation Is Starting to Take Shape
Week 1 on the live account is done. Weekly realized P&L: +₩40,658. But once unrealized P&L is included, the closing balance was ₩13,941 lower than where the week started. Taken at face value, it’s hard to call this a clear win yet.
But the metric I was actually watching this week wasn’t the return.
What This Week Was Really About
After switching BUY_DRY_RUN to False, the system placed its first real orders on its own. It scanned each morning, picked stocks that passed the RS filter, submitted the orders, and when a trailing stop fired in the afternoon, the sell went through. The key trading decisions and order submissions were handled within the system’s own flow — no operational downtime, no API errors. Zero stop-losses out of 7 trades.
That’s the core result for this week.
What the Trailing Wins Showed
Samsung Electronics, NAVER, KB Financial — all three exits were positive. Samsung fired on a gap-down, so the gain was modest, but it didn’t go negative. The trailing stop’s approach — allow a defined range from the peak before selling — worked as intended in live conditions.
Between 4/20 and 4/22, stop-losses locked in –₩61,800. Trailing P&L through this week has reached +₩51,658. Recovery rate: 83.6%. The structure where trailing gains gradually follow stop-loss losses is actually forming.
What Cash Recycling Means
On the morning of 4/30, available buying power was around ₩70,000–₩80,000. Even when RS-qualified candidates appeared in the scan, there wasn’t enough capital to buy. Only after KB Financial was sold at 13:50 did cash recover enough to enter LG Electronics at the 14:00 scan.
At first this looked like a structural constraint. Looking at it again, it’s the right design. The system needs to empty a position before it can take the next opportunity. It doesn’t force entries. When an existing position exits with a gain, the next candidate fills that slot. Rotation has begun.
What the RS Filter Changed
Samsung Electronics was RS-qualified on 4/29 but fell out of the filter on 4/30. The KOSPI ETF itself rallied hard, making Samsung’s relative strength comparatively lower. The system detected this automatically and didn’t buy.
The stronger the market, the stronger a stock has to be to survive the filter. That’s what the RS filter does. This week was the first live confirmation.
What’s Still Open
The web export error was supposedly fixed, but it reappeared on 4/30. Log visibility has improved, but the data export side is still unstable.
SK Hynix ranks first by RS but its share price of around ₩1,300,000 exceeds the per-trade budget. The question is whether to leave it in the universe anyway, or exclude it from the start. Shinhan Financial keeps getting blocked at MA20 deviation of 1.7–1.9%. Whether to adjust the parameter or hold the current threshold is a decision for this week.
The Operating Principle Going Forward
What the system is doing right now is validating the structure. Growing the returns comes after that. Error-free execution, trailing stops working as designed, position rotation repeating — I need a few more weeks of confirmation before the next phase can even be discussed.
This week was the first of those weeks.
This is a personal automated trading system operations log. Not investment advice.