Reading Earnings Scans — Pre-Earnings Options Strategy Ranker

The four strategies in one paragraph each

Earnings Rush — long IV before the report

IV typically expands into earnings as uncertainty builds, then crushes the morning after the report. Rush plays the expansion: enter a long ATM straddle ~10 trading days pre-EE, close BEFORE the announcement. Wins on vega regardless of which way the underlying moves; loses when IV doesn't rise (already pricing in the event) or theta exceeds vega. The V3.4 backtest uses a longer-dated expiry (first Friday ≥21 days post-EE) because short weeklies have tiny vega and theta wipes out any IV gain. Exit is the close BEFORE the announcement: EE-day close for AMC reporters, prior-day close for BMO.

Iron Condor — short premium through the event

Sell an OTM put spread + an OTM call spread, held through the report. Wins on IV crush plus stock staying inside the inner strikes. The V1 score scales with (implied move − historical |move|) — bigger gap = more juice to harvest, less risk of the wings being tested. The V3.2 backtest replaces that guess with: did this exact 1.0×/1.5× condor structure actually make money across the last 6 earnings? Win %, Avg ROI, sparkline — straight from Polygon contract aggregates.

Straddle — long premium through the event

Buy ATM call + ATM put, hold through the report, exit after the IV crush. Wins when the stock moves further than the straddle cost. The V1 score is the inverse of Condor: high when historical |move| has reliably EXCEEDED implied move. The V3.1 backtest checks: simulating this exact ATM-straddle trade on the last 6 earnings cycles using real Polygon prices, what was the actual Win % and Avg ROI?

Breakout — directional bet on follow-through

Some stocks have a strong post-EE directional bias — they consistently rip in one direction after earnings (or consistently dump). Breakout buys a single-leg ATM call or put pre-EE aligned with that bias. The V3.3 backtest is rolling-window — at each past cycle, direction is decided by the mean of OLDER cycles only (no look-ahead). That keeps the simulation honest: if a ticker truly has a persistent directional bias, the rolling signal catches it; if the bias is just noise from the latest few quarters, the backtest will show low win rate or negative ROI and rank low.

How sort order works

On Straddle, Condor, and Breakout tabs, rows are sorted by:

1. Confidence tier descending — STRONG before WEAK before THIN before none. A 60% win rate on 5 cycles ranks above a 100% win rate on 1 cycle, every time.
2. Avg ROI descending — within tier, the trade that made more money historically ranks first.
3. V1 heuristic score descending — final tie-breaker when backtest data is sparse.

On Rush and Breakout tabs (pending V3.3 / V3.4), rows are filtered to V1 score ≥ 50 and sorted by score descending. The ≥50 floor exists because the V1 heuristic is noisy at the low end and showing every 0-49 row would drown the signal.

The banner state machine

The top of every backtested tab shows one of three banners:

• Emerald — ✓ N STRONG PICKS: N tickers cleared the ≥4-cycle bar. Those are the actionable rows — focus there.
• Amber — ⚠ NO STRONG PICKS:No row hit ≥4 cycles, but at least one returned 2-3 (WEAK). Treat the WEAK rows as a watchlist worth monitoring; don't commit capital based on a 2-3 cycle sample alone.
• Gray — NO QUALIFIED CANDIDATES:Zero rows have enough priceable historical option cycles to produce a reliable backtest. This is common in slow earnings weeks or for weeks dominated by recent IPOs. The absence of edge is itself information — there's nothing this scanner finds tradeable, so go work on something else.

Common confusion: V1 score vs V3 backtest

Important distinction: the 0-100 Strategy Score is V1 heuristic — a smart guess based on comparing implied move to historical |move|. The Win % / Avg ROI / sparkline cell on Straddle and Condor tabs is V3 backtest — actual P&L from simulating the strategy on real historical Polygon prices. They can disagree:

• A V1 score of 76 with WEAK or THIN backtest tier means “the setup looks favorable on paper, but we don't have enough historical data to confirm it actually works for this ticker.”
• A V1 score of 30 with STRONG +40% Avg ROI backtest means the heuristic underrates this setup — the real history says the trade has worked. Trust the backtest.
• When both agree (high V1 + STRONG positive backtest), that's the highest-conviction pick of the week.

Mechanics under the hood

The backtest engine for one ticker:

1. Fetch the last 6 earnings dates from Finnhub.
2. Fetch underlying daily bars for the full date range (one Polygon call per ticker).
3. Fetch the current option chain once to determine the ticker's expiry cadence (weekly / monthly / both).
4. For each past earnings cycle:
   • Compute entry date (4 trading days pre-EE) and exit date (1 trading day post-EE, accounting for BMO/AMC timing).
   • Pick strikes — ATM-rounded for Straddle, 1.0×/1.5× implied move for Condor (sized by current implied move applied to entry-day spot).
   • Build OPRA contract tickers (e.g. O:AAPL250117C00150000).
   • Fetch per-contract aggregates for the entry-to-exit window (Polygon /v2/aggs/ticker/O:.../range/1/day/...).
   • Compute entry price, exit price, per-cycle P&L, ROI.
5. Aggregate the cycles into Win %, Avg ROI, cyclesUsed (cycles that produced priceable bars), totalCycles (cycles attempted).

Cycles can fail to price for legit reasons: thinly-traded historical strikes, weekly expiries that weren't listed back then, holiday gaps, contract tickers that don't exist for low-priced names with wide strike grids. Those failures show up as the cyclesUsed < totalCycles gap and drive the confidence tier down accordingly.

What to do with this every week

1. Sunday night / Monday morning: open the page, check the banner state on Straddle and Condor.
2. If STRONG picks exist: open the top 2-3 rows. Read the Win %, Avg ROI, and sparkline. Cross-check the V1 rationale. Hit BUILD to drop the structure into Risk Graph.
3. If only WEAK / THIN / none:skip the backtested tabs. Glance at Rush and Breakout for directional ideas. If nothing there either, go work on something else — there's no edge this week.
4. Cross-reference: Options Edge anomalies + GEX dealer-positioning often line up with the same names this scanner flags. A trade with backtest confirmation + IV anomaly + GEX alignment is much higher conviction than backtest alone.