Finacks — 2026-04-24 [PAPER]

Market Context

Tech momentum dominated the session. Broad indices and leveraged tech ETFs (FNGU +26.6%, UPRO +15.0%) signaled strong bullish appetite, though individual stock gains were more modest. The environment rewarded both momentum holds and selective entry discipline.

Strategy & Decisions

The bot cycled through five strategy variants centered on the same core thesis: hold winning positions while selectively adding to momentum plays in META, NVDA, and broad tech exposure. All variants shared this bias—no contrarian or short-bias strategies were deployed. The approach relied on RSI and MACD as primary momentum filters. Claude generated no explicit market assessments during the day, which limited real-time reasoning documentation.

Trades

AMD: -2.1% ($-7.28) closed via claude_sell signal. The exit rationale referenced AMD's +11.57% intraday gain paired with RSI=97.3 and MACD histogram=8.62, signaling overbought conditions. This was a reasonable mean-reversion exit but proved slightly premature—the position still closed the day +0.4%, suggesting the sell may have occurred mid-session before additional upside.

Positions Carried Overnight

11 positions held at close:

• Large winners: FNGU (+26.6%), UPRO (+15.0%)—leveraged tech bets that captured the day's momentum
• Core holdings: AAPL, AMZN, GOOGL, MSFT (all +1.3% to +2.1%)—broad tech exposure
• Smaller gainers: NVDA, ISRG, TQQQ, AMD (+0.4% to +0.7%)
• Laggard: SOXL (-0.4%), though the bot noted extreme RSI=100 and acknowledged it's not currently held in the standard portfolio

Performance

Day P&L: +$201.36 (+0.19%) on $104K equity—a modest but positive close. The 26 cycles run reflect heavy optimization iteration, yet the rolling 30-day stats (28% win rate, -$21.27 P&L across 86 trades) tell a harder story: the bot is underwater month-to-date despite today's gain. Win rate near 28% is concerning and suggests either poor entry discipline or unfavorable market conditions for the current strategy set.

Takeaway

Momentum filters alone aren't sufficient. The AMD exit highlighted the tension: RSI=97 looks extreme, but the market kept climbing. Leveraged ETFs captured outsized gains (FNGU, UPRO), but the 28% win rate across 86 trades suggests the bot is overtrading noise and cutting winners too early. Consider: reducing cycle frequency, raising the bar for new entries, and letting winning positions run longer before taking overbought signals at face value.