Strategy Deep-Dive

Long Straddle Backtest: 100 Earnings Events, Brutal Average

Buying ATM straddles into the top 25 mega-cap earnings reports over 4 quarters. The averages are ugly — but the tails are interesting.

Simulated data for display. Illustrative narrative — not a verified live backtest.

Earnings Events
100
top 25 mega-caps, 4Q
Win Rate
38%
below break-even
Net P&L
-$1,800
per contract sample
Best Single Trade
+$1,460
NVDA Q3 surprise

The Test Setup

Universe: Top 25 S&P 500 mega-caps by market cap. Earnings events from Q1 2024 through Q4 2024 inclusive.

Entry: ATM straddle (call + put at the strike closest to spot) bought at the close on the day before earnings release.

Exit: Close at the open the morning after earnings, before 10am ET.

Expiration: Nearest weekly that's at least 7 DTE from entry, to balance IV expansion vs theta drag.

Sizing: One contract per event. No position scaling.

The Headline Result: IV Crush Wins

Across 100 earnings events:

Bucket Count % of Sample Avg P&L
Big winners (>100% gain)1212%+$760
Modest winners2626%+$220
Small losers (-20% to 0)3535%-$120
Big losers (over -50%)2727%-$380

62% of trades lose money. The 12% big-winner bucket can't quite offset the 62% combined loser buckets. Net result is a slow grind down for systematic earnings straddle buyers.

By IV Rank at Entry

The single most useful filter found in the backtest:

IV Rank at Entry Trades Win Rate Avg P&L
Under 201861%+$240
20–402846%-$80
40–603131%-$285
Over 602318%-$485

The IV-rank rule: only buy straddles into earnings when IV rank is below 20. In that bucket the strategy was net profitable. Every other IV regime was a money-loser. Most earnings events have IV rank 40+ at entry because options markets pre-emptively bid up vol — which is precisely why systematic earnings straddle buying loses on average.

Anatomy of the Best Trade: NVDA Q3 2024

Entered NVDA $135 straddle on the day before Q3 earnings at $9.20 debit. NVDA opened up 22% the next morning at $165, reflecting a massive top-line beat and AI capex guidance. The call leg was worth $30+ at the open. Closed for $23.80 credit, +$1,460 per contract.

The signal that this trade was different: IV rank was only 18 going into the report, well below the typical 50+ for NVDA earnings. The market was complacent — underpricing the realized move that subsequently came in dramatically above the implied move.

These "vol underpricing into binary event" setups are the only structural edge in long earnings straddles. They show up roughly 12% of the time. The rest of the cycles are IV-crush losers.

Common Mistakes

  • Buying into elevated IV rank. The market is already pricing the move; you're just paying for IV crush.
  • Going too far OTM with the strike. ATM straddles capture vol expansion most efficiently. OTM strikes are cheaper but need bigger moves.
  • Holding through multiple earnings cycles. Theta eats the position; close after the binary event.
  • Treating every earnings event as an opportunity. Selectivity is the only edge.
  • Position sizing on expected return rather than tail probability. Most cycles lose; survive long enough to catch the rare big winner.

Five Takeaways

  1. Systematic earnings straddle buying is a losing strategy. 62% of trades lose; the rare big winners don't quite compensate.
  2. Only buy when IV rank is under 20. The only profitable bucket in the entire backtest.
  3. Selectivity beats coverage. 12 trades a quarter beats 100 trades a year on net P&L.
  4. Selling straddles is more profitable on average — with brutal tail risk. Most short-straddle traders eventually graduate to defined-risk structures after a blow-up.
  5. Look for vol mispricing, not earnings. The edge is in finding names where IV is structurally low relative to historical realized moves — not in systematically trading every report.

Find vol mispricing in real time

Our IV rank scanner highlights names with low IV ahead of binary events — the rare setups where long straddles produce positive expectancy.

Related Reading

Backtest narrative is illustrative — built from typical earnings volatility patterns, not from live broker fills. Past performance, simulated or real, does not predict future results. See methodology.

7 days free, cancel anytime No charge if you cancel
Start trial →