farbzilla— how it works

For each of ~200 large-cap US equities we compute five signals derived from options-market activity — things like deep-OTM call volume relative to its 60-day average, the average moneyness of call trades, and total overall moneyness skew. These capture moments when options flow is unusually directional.

Each day we rank every name cross-sectionally on each signal, average the percentile ranks into one composite score, and go long the top 10, short the bottom 10.

Signals only fire when the broad market is calm and trending: VIX < 20 and SPY above its 200-day SMA. Stepping out during high-VIX or below-SMA regimes nearly doubled Sharpe and cut drawdown in backtests.

Positions are held ~10 trading days, then closed unconditionally. Sizing is equal dollar — 5% of starting capital per leg, so 20 legs ≈ 100% gross exposure (long/short, roughly market-neutral).

The live account is a Tradier paper-trading sandbox. Every order is recorded here and the equity curve you see above is real day-by-day P&L on that account — no simulation on top of the live data.

• Sharpe ≈ 2.58 after realistic 25-bps round-trip costs
• Max drawdown ≈ -7.3%
• Pre-cost alpha positive on market-neutral (long-short residual) tests
• Survives cost sensitivity up to ~50 bps

These are backtest numbers, not guarantees. Live performance can and will differ.

• Not financial advice. It's a free hobby dashboard.
• Not options trading — the strategy uses options flow as signal but trades the underlying stocks.
• Not a subscription or paid Discord. Follow along if you want; don't copy blindly.