Skip to content
VolSurface Desk
Directory

Index Dispersion Correlation Desk

Pre-Earnings Implied Jump Extraction (σ_earn): Stripping Event Variance

Four times a year, single-stock weekly implied volatilities surge to 65%–110% ahead of quarterly earnings reports while SPX implied volatility barely budges. Derive the exact additive total-variance extraction formula for σ_earn.

By Dr. Alistair Finch, CQF, FRMReview not recorded · Updated
Pre-Earnings Implied Jump Extraction (σ_earn): Stripping Event Variance — VolSurface Desk 2026 quantitative & supervisory data card
VolSurface Desk (Implied Correlation, Index Dispersion Baskets & Earnings Vol Crush): Two-tenor additive total-variance stripping for single-stock earnings jumps and index aggregation.Source: VolSurface Desk Institutional Research DeskLicense: CC-BY-4.0

During the third week of January, April, July, and October, single-stock weekly options on companies like Netflix, Meta, Tesla, and Nvidia routinely trade at 70% to 115% annualized implied volatility, while 5-day Cboe SPX Weeklys trade calmly at 13.5%. For an institutional Index Dispersion trader, guessing whether 82% Friday implied volatility on Meta is "expensive" or "cheap" without stripping out baseline diffusive variance is impossible. You must extract the exact Implied Earnings Jump Standard Deviation (σ_earn).

Within our Implied Correlation, Index Dispersion Baskets & Earnings Vol Crush silo and Dispersion Desk Manual, this note derives the institutional Two-Tenor Forward Variance Stripper and proves how individual corporate jumps aggregate into S&P 500 and Nasdaq-100 index chains.

Exact Two-Tenor Additive Total-Variance Extraction of σ_earn#

Work strictly in business/trading days (Nbus = 252 days/year, excluding weekends and exchange holidays, or weighting non-event trading days at 1.0 and weekends at 0.15). Let expiration 1 have D1 effective trading days remaining and expiration 2 have D2 > D1 effective trading days remaining, where both D1 and D2 occur AFTER the scheduled earnings release (and no second earnings event falls between D1 and D2). Total cumulative variance W(Dk) = (Dk / 252) · σATM2(Dk) decomposes linearly:

Formula or model from the articleNot independently verified
v_diff = [ W(D2) − W(D1) ] / (D2 − D1) , σ_earn = √[ W(D1) − D1 · v_diff ] , E[|Jump|] = √(2/π) · σ_earn ≈ 0.7979 · σ_earn
  • Where W(Dk) = (Dk / 252) · σk2 is cumulative dimensionless variance to expiry k, vdiff is daily diffusive variance, and σdiffann = √(252 · v_diff) is ex-earnings annualized volatility!
  • Worked Live Example — Nvidia (NVDA) Ahead of Wednesday AMC Earnings (Evaluated Monday Noon: D1 = 4.5 trading days to Friday expiry at σ1 = 78.40%; D2 = 14.5 trading days to 2-week expiry at σ2 = 51.90%):
  • 1) Cumulative Variances: W(4.5) = (4.5 / 252) × 0.78402 = 0.010976; W(14.5) = (14.5 / 252) × 0.51902 = 0.015501.
  • 2) Daily Diffusive Variance: vdiff = (0.015501 − 0.010976) / (14.5 − 4.5) = 0.0004525/day ⇒ Ex-Earnings Annualized Vol σdiff = √(252 × 0.0004525) = 33.77%!
  • 3) Pure Earnings Jump: σ_earn = √[ 0.010976 − 4.5 × 0.0004525 ] = √[ 0.008940 ] = ±9.455% (1σ Gaussian StdDev) ⇒ Implied Mean Absolute Gap E[|Jump|] = 0.7979 × 9.455% = ±7.54%!

Traders verify ATM straddle dollar premiums and post-print implied volatility collapse targets directly in our Black-Scholes Calculator, Implied Volatility Calculator, and 0DTE Expected Move Calculator.

Pre-Earnings Implied Jump Extraction & Dispersion Screening Pipeline#

Four-step quantitative workflow to extract single-stock sigma_earn and score historical earnings dispersion value
Four-step quantitative workflow to extract single-stock sigma_earn and score historical earnings dispersion valueVerified Exhibit

Why Earnings Jumps Diversify Quadratically Out of S&P 500 Variance!#

Here is the mathematical engine that powers quarterly dispersion desks: suppose two companies i and j report purely idiosyncratic earnings surprises εi and εj (with zero cross-company earnings surprise correlation, Cov(εi, εj) ≈ 0, aside from shared sector read-throughs like AMD/NVDA or META/GOOGL). How much jump variance enters the S&P 500 index?

  • Constituent Basket Leg: Receives the linear weighted sum ∑ wi · σearn,i2!
  • S&P 500 Index Leg: Receives only the quadratically attenuated sum ∑ wi2 · σearn,i2 + ∑{i≠j ∈ same sub-sector} wi wj ρreadthrough σearn,i σearn,j!

Even for Nvidia at a massive w_NVDA = 6.8% SPX weight, wNVDA2 = 0.004624! A ±9.46% single-stock earnings jump (σearn2 = 0.00894) contributes just 0.004624 × 0.00894 = 0.0000413 to cumulative SPX variance—barely ±0.64% of one-time SPX move! For a 0.8% weight constituent (like Netflix or Caterpillar), w_i2 = 0.000064—its ±10% earnings jump is 99.99% diversified away inside SPX!

Representative Earnings Jump Profiles Across S&P 500 Tiers#

Quantitative Distribution · Typical SPX Weight (w_i)
Typical SPX Weight (w_i)Ex-Earnings Diffusive IV
01AI Semiconductor Bellwether…2.2% – 7.0%02Mega-Cap Platform (MSFT, AA…4.0% – 7.3%03High-Beta Consumer / Stream…0.9% – 2.1%04Large-Cap Financials / Heal…1.1% – 1.8%
Constituent ArchetypeTypical SPX Weight (w_i)Ex-Earnings Diffusive IVPre-Earnings Weekly IV (4.5d)Extracted 1σ Jump (σ_earn)Index Variance Leak (wi2 · σearn2)
AI Semiconductor Bellwether (NVDA, AVGO)2.2% – 7.0%32% – 38%72% – 88%±8.5% to ±10.5%Noticeable (+0.4v to +0.8v on SPX; +1.8v on NDX)
Mega-Cap Platform (MSFT, AAPL, GOOGL, AMZN)4.0% – 7.3%20% – 26%42% – 58%±4.2% to ±6.8%Moderate (+0.2v to +0.5v when clustered)
High-Beta Consumer / Streaming (NFLX, TSLA)0.9% – 2.1%34% – 48%78% – 105%±9.2% to ±12.4%Tiny (< 0.08v on SPX; pure dispersion fuel!)
Large-Cap Financials / Healthcare (JPM, UNH, LLY)1.1% – 1.8%17% – 24%34% – 48%±3.6% to ±5.8%Negligible (< 0.03v on SPX)

Implied Volatility Term-Structure Kink Ahead of a ±9.46% Earnings Print#

Term-structure volatility decay from 4.5-day weekly expiry out to 90-day quarterly expiry ahead of a ±9.46% earnings jump
Term-structure volatility decay from 4.5-day weekly expiry out to 90-day quarterly expiry ahead of a ±9.46% earnings jumpVerified Exhibit

Once earnings hit the tape at 16:05 ET and weekly implied volatility collapses by 40+ vol points the next morning, desks transition immediately to our sibling playbook on Post-Earnings Volatility Crush & PEAD Diagonal Spreads. Compare live parameters in Dispersion Correlation-Spike Tail Risk & Crash Hedging.

Frequently asked questions

Why must you divide annualized implied variance by 252 trading days (not 365 calendar days) when extracting σ_earn?

Equity volatility accrues overwhelmingly during regular and extended trading sessions rather than quiet weekends! Dividing annualized σ2 by 365 understates daily diffusive variance by 31% (252/365 = 0.69), artificially inflating your extracted earnings jump σ_earn on Tuesday-through-Thursday expirations.

How does the Options Market Maker ATM Straddle rule (±0.80 × Straddle %) compare to exact variance-extracted σ_earn?

For an option expiring the Friday immediately after a Thursday after-market-close (AMC) earnings print (T1 ≈ 1 trading day left), Total Straddle ≈ √(2/π) · S · √(σ_diff2 + σ_earn2) ≈ 0.7979 · S · σ_earn! Thus Straddle % / 0.7979 gives 1σ standard deviation, while raw Straddle % (~0.80σ) equals the Expected Mean Absolute Deviation E[|ΔS/S|]!

Why can two weekly expiries T1 and T2 occasionally produce a negative extracted σ_earn2—and what causes it?

If T1 spans earnings while T2 ALSO spans a major macro FOMC/CPI event or a second product launch (like Apple WWDC or Nvidia GTC), then diffusive/macro variance between T1 and T2 exceeds baseline σ_diff2, biasing (T2σ22 − T1σ12)/(T2−T1) upward! Always subtract known FOMC/CPI macro variance weights from SPX/QQQ first.

How much does Peak Tech Earnings Week lift Nasdaq-100 (NDX) and S&P 500 (SPX) weekly implied volatility?

Suppose MSFT (7.0% SPX wt, σ_earn = ±5.2%), AAPL (7.2% wt, ±4.5%), META (2.8% wt, ±8.5%), AMZN (4.1% wt, ±6.8%), and GOOGL (4.2% wt, ±6.0%) all report in one week: ∑ w_i2·σ_earn,i2 ≈ 0.000435, which adds √(252/5 × 0.000435) ≈ +4.7% annualized variance—lifting 5-day SPX IV from 13.0% to ~13.8%, and NDX IV (where weights are 1.5× larger!) from 17.0% to ~19.4%!

Publication record

Published
November 13, 2026
Updated
November 14, 2026
Author
Dr. Alistair Finch, CQF, FRM
Review
Review not recorded
Automated fact check
Passed in the stored pipeline record
Source URLs
3 recorded

AI assistance recorded: anthropic/claude-opus-4.6. This field does not establish human review or source verification.

Academic & Institutional Attribution

Cite This Research & Embed Live Data Card

Open Attribution License (CC BY 4.0)

Standard academic & institutional research paper citation format.

Dr. Alistair Finch, CQF, FRM. (2026, November 14). Pre-Earnings Implied Jump Extraction (σ_earn): Stripping Event Variance. VolSurface Desk. https://volsurfacedesk.com/desk-notes/pre-earnings-implied-jump-variance-extraction-sigma-earn
Explore Index Dispersion Correlation Desk Hub (20) →

VolSurface Desk Data, Benchmarks & Contributor Records

Pre-Earnings Implied Jump Variance Extraction (sigma_earn) | VolSurface Desk